A lot of vibrational absorption bands that absorb microwave frequencies are due to the variation in local sites of atoms in a polymer. It's not that PTFE or HDPE don't have resonance, it's that because they're very regular polymers, the resonances tend to be the same at each site rather than varying from site to site. Because there are sharp resonance frequencies, most of the microwave spectrum is unaffected by the resonances. For example, in PTFE or HDPE, the C-F and C-H bonds respective are identical because of the repetitiveness of the polymer chain, and so each C-F and C-H bond looks the same relative to its neighbors, except perhaps the few bonds on the end of the chain, and perhaps where there are crosslinks if there are crosslinks between the chains. Therefore they all behave similarly and have the same resonances. For this reason regularity of the polymer is very important because this tends to cluster the resonances into a few sharp bands that can be avoided.
The vibrational modes of really any molecule (the vibronic energies) are not really microwave frequencies. Rather, the microwave absorptions and emission (scattering) operates via "domains" in which the contribution of many molecular subunits add up to a collective dipole. Its also possible to excite higher order moments (octopoles) through induced "polarizability" but that is probably not a significant contributor for mostly aliphatic hydrocarbon materials like plastic. For people who are familiar with their basic chemistry laboratory spectroscopy, the characteristic frequencies of hydrocarbons are in the ~2950 and ~1490 "wavenumbers" ("inverse centimeters") corresponding to ~3.4 micron and ~6.7 micron respectively (wavenumbers scale like frequency rather than like wavelength in terms of energy) Incidentally, this is also seen in home microwave ovens. The water molecules vibrational excitaitons are nowhere near the ~2.45 ghz microwave frequency radiation that the magnetrons are tuned to create. Rather, collective phonon like oscillations of clusters of water molecules are the underlying molecular dynamical process that is driven by the external field. That value, 2.45 ghz, is also a combination of the material properties of water/food and practical optimization that takes into consideration the penetration depth of the radiation and considers the most likely temperature range of the food (~5 C "just out of the fridge" to 100 C atmospheric pressure water boiling point). The excitation of clusters of water molecules at this energy is also useful to understand why Ice is difficult to heat. The molecules are constrained individually in their little ice domains, and even more so when one considers the collective dynamics of many molecules together (a topic that is likely even more complicated given the various forms of ice that can form and the differences in how those different forms interact with radiation far below resonant/excitation energies) and, additionally, as these are solid state compounds, *ro*vibronic excitations are also not really a thing since you can't excite states in the rotational manifold of a molecule that is frozen in solid state in the gas phase, rotational excitations are more in line with "microwave" (or terahertz) energies.
To heat up a couple of dinner plates in a microwave I put some water in the bottom plate and place the other plate upside-down on top and heat for a couple on minutes. This works well. So to heat a couple of bowls I did the same, I placed some water in the bottom bowl, but instead of putting the top bowl upside-down, I nested it the right-side up. This created a curved lens of water in the bottom bowl that focused the microwave energy down onto the plastic spider bracket supporting the glass turntable. The spider bracket caught on fire! Once the plastic in the spider bracket started the carbonise, this seemed to concentrate the microwave energy in the carbon and a runaway effect ensued. Lesson… don’t put nested bowls separated by water in a microwave!
@@MaakaSakuranbo I have some plates like that, they must have some polar minerals that are causing conversion of RF energy into heat. Could be the glaze or the bulk material
Refraction is REALLY weird. Even if you don't go into the fine detail of the quantum-mechanical workings, the mechanism of refraction at microwave frequencies at the molecular level is not what most folks think. My little aside about Ibn Sahl and medieval Persian rather missed the fact that the image in annotated in Arabic. My late wife Caroline would have spotted that in an instant and corrected me.
Also, since the the refractive index is also proportional to the square root of the relative permeability, a similar effect will happen with the magnetic moments as well as the dipole moments for the general case, although your 3d printed lenses are obviously not paramagnetic. It was good to open up Bleaney & Bleaney again and reread the chapter on dielectrics....
By the middle ages, I'd have thought the Persians were writing Persian in Arabic script anyway... But if I remember my Jim Al Khalilli correctly, stuff to be discussed in the majlis would have all been in best schoolroom Arabic much like medieval Christian science was all written in best schoolroom Latin.
@@Mike-H_UK There aren't many magnetic materials around that will work above a few hundred MHz. Rogers MAGTREX 555 is usable up to 500MHz or so. It has the same numeric value for relative permeability and permittivity, which has some interesting applications. I just wish it would work at 3GHz or so
I feel like watching Alec from Technology Connections and that's a compliment! I don't know why am I watching this and I understand less than half of what you're saying and still this is fun to watch and pretty interesting! ... Well, maybe I understand a bit more but the puns are great!
thank you for this video. i really appreciate your approach, using convenient simplifications but calling them out as you do so. i found it easier to focus on the simplifications, but it's great to know where they are. i also note your use of words like "apparent", to carefully work around the simplifications. this is a very good approach for bringing people into the subject.
@@adfaklsdjf I've suffered too often over the last 60 years from over simplification and under explanation of complexity. I'm not very good at this sort of thing but I think it's important to explain when I am making ludicrous simplifications or scrubbing over things that would take me three hours to explain. Thanks for being an excellent viewer. Top marks!
Thanks Neil. I wish that you could have been around 45 years ago when I was making and testing microwave transistors. I doubt that most people around me at the time really understood how some of the test equipment really worked.
in 1973 and 74 I had a summer job working at AEI semiconductors in Lincoln. In my second year I was working on some tests of overlay transistors as they called them at the time, although I can't remember what frequency we were working at. The first year I worked mainly with varactor diodes. My auntie Val worked there doing die bonding and other extremely precise work with transistors and later with MMICs. I was only 15 years old when I started and had to get special permission to be allowed to work. Then I went off to university and ended up messing about with computers, despite having zero interest in them. Even now I'm still working part time with beastly computers after 50 years. I sometimes wonder how different life might have been if I'd have carried on with electronics and radio. I suspect that it would've spoiled my interest in the subject areas in the same way that being paid to mess with computers, networking and communications wiped out any hobby interest in those areas.
I'm an electrical engineer, and not an RF engineer, but I've always found this field fascinating. It's why I watched the original. I do have some basic RF understanding though. Also, programming is interesting in much the same way that there are people interested in linguistics or specific spoken languages. It's about working together on something whether it be human-machine communication, or human-human communication. It's about synergy and the constant drive to improve, adapt and overcome obstacles that are apparent boundaries in that communication. Much the same way you seem to have an interest in RF tech, and are excited about new breakthroughs in the medium. The same can be done with communication.
I have a very jaundiced view of programmers as a result of spending all of this century so far dealing with poorly-written code that's caused security defects. I only see broken code, I'm sure there must be non-broken code out there, but I never get to see any of it in my day job!
Neil, your description of atomic interaction with an electromagnetic wave is the most cohesive explanation of refraction I've ever heard. Nice to see Aimee make an appearance, as well. Thanks for another very interesting video.
I'm sad that I had to chop out the quantum explanation and the bits about Polarons, but perhaps that's something for my other channel. I would have liked to create a model of a good dielectric based on Lorentz oscillators, but life is too short, I have a lot of machining to catch up with!
Medical X-ray machines that use the peeling of adhesive tape from a roll are really interesting. The came out around 15-20 years ago. It was a way to make an X-ray without a power supply. The breaking of the electron bonds created X-rays that interacted with the X-ray sensitive film.
I remember trying to get images of the fluorescent flashes from sellotape being peeled back in the 90s. I could see them in a dark room, but never managed to get useable images. Interesting!
Nice lesson, but eventually this approach got me thinking about Fresnel lenses. In other words, a Fresnel lens for radio waves. I'm not convinced that directivity can replace the meaning of gain, as you said you consider the same. Gain needs a reference (isotropic, dipole, etc.), but directivity and gain aren't necessarily related, they are two different things. It's possible to have great directivity and still produce negative gain.
I don't think it's useful to talk about gain with passive aerial systems, as they don't really have anything other than loss, but I was being deliberately provocative there. You are 100% correct of course, Directivity and Loss are both important for antenna systems. I'm working right now on some 3D printed Fresnel Zone Plates and lenses using some special dielectric filaments, I hope to get some more videos out before the end of 2023. My life has become extremely complicated over the last six months and my video work has gone onto the back burner for a while. Fresnel Zone Plates about 180 mm diameter have useful gain at frequencies over 47 GHz, but I'm working on some large versions that are printed in segments, with anti-reflection chokes on the flat surfaces, but getting the supports right is proving rather difficult. I may need to look at at alternative solution for the AR surfaces unless I can get clean enough supports. I might need to change to a soluble support filament to get the best possible results.
The light doesn’t appear to slow down, it really does slow down. The “speed of light” is more properly called the “speed of causality”, which is the maximum relative speed that any pair of things can have, but that does not mean light waves must travel that fast in matter. Source: I work with ultrafast pulsed lasers for my research, so the slowing down in matter is really significant for us
Interesting! Does it slow down though? Sure the pulses take longer to get from one side of a piece of material than they would in a vacuum, so at a macroscopic level, they certainly appear to be travelling slowly, but what's the molecular-level or quantum mechanism for the slowing? Don't the laser photons interact with the electron clouds and get re-radiated by coherent forward scatter in a similar way to what happens with molecular bond coherent forward scatter at microwave frequencies? The coherent scattering has an inherent phase lag because the electron charge cloud and nucleus have rest mass. They are forced to move apart or together slightly by the electric field of the incident EM waves. They then radiate that energy as they are moving charges, but with a phase lag. Then the superposition of all of those interacting electron cloud/nucleus oscillations sums to an output that's delayed in time? Talking classical here rather than quantum field theory, but what is the mechanism that can slow light down without the presence of gravitational gradients strong enough to cause relativistic effects? What's the mean free path of your femtosecond pulses in a solid medium compared with the size of the electron clouds and inter-molecular spacing?
@@amirsadeghi9888 Metamaterials can make it appear that the speed of waves through a material is superluminal, but it's usually done using resonant structures and the effect only works in bulk. You can't get any information to travel faster than light, so causality isn't affected, it's more about a negative refractive index, where lenses bend light the "wrong" way to make sort of invisibility cloaks. Kind of...
@@MachiningandMicrowaves wow thanks for the response, very interesting, but what if I want it to affect causalty,lol. Metamaterial and gaseous lenses and laser couplin can create negative reflective index and make light bent the "wrong" way out or appear it travels faster than light, but what about quantum tunneling and "spooky actions at distance" which aparantly travels faster than light...? there must be a way to go superluminal without blowing up the universe, lol. perhaps an ai with a quantum computing operating system can help solve going warpspeed
@@amirsadeghi9888 quantum entanglement does work faster than light, but there isn't any extractable information conveyed, so the speed of causality isn't violated. There really isn't a way to go superluminal. If you really want to go warpspeed you'll have to bend spacetime in order to reduce the distance instead of trying to increase your speed.
Even the 23:xx version made the previous printed lens video(s) make INFINITELY more sense. I was trying to make sense of various waveforms interacting at 1:1 scale with the crazy organic-ish shape, but duh. Of course you can just vary the characteristic wall thickness, so long as the unit cell is small compared to the wavelength. I'm looking forward to many great videos in the future from this channel.
I've had fun playing with microwave technology for more than fifty years and it's still endlessly fascinating. So much more interesting than my Day Job as a cybersecurity architect and occasional network and firewall engineer and forensics investigator. Materials science is endlessly interesting. I did a degree in Astrophysics a few years ago when the grandchildren were complaining about how us Old People had it SO easy, whereas they had to do school and University exams. Perhaps I should go study a bit of condensed-matter Physics next.
Could you do a video on how negative db is measured? Example from satellites that are far away they often talk about signal strength in -160db. How do you measure a negative signal? 🤔
dB is a measure of logarithmic ratio. Signal strength is usually measured relative to one milliwatt as dBm, so if a signal is one nanowatt, it is a millionth of a milliwatt, so as a power ratio it's 10^-6. Convert that to a base-10 log and you get -6, now to convert to decibels, multiply by 10, that gets you -60 dBm. A picowatt is -90 dBm, a femtowatt is -120 dBm, so the power represented by -160 dBm is about 100 zeptowatts. Thermal noise at room temperature in a resistive source is about -174 dBm/Hz, so it's close to impossible to receive a -160 dBm signal at GHz frequencies because of phase noise and atmospheric scintillation, even from a distant satellite. However, if the receiver is cooled and the antenna is pointing at cold sky, you can do better than Boltzmann's constant would imply at room temperature, so -160 dBm isn't impossible, so long as the bandwidth (and data rate) are low. The path loss to the Moon and back at 10 GHz, along with the 7% reflectance of the lunar regolith, is around 289 dB, so with a 20 watt transmitter (43 dBm) and 50 dB gain antenna, I can send out 93 dBm. A receiving station with an antenna gain of 50 dB also reduces the path loss, so I pick up 143 dB. I should expect a signal of about 143 - 289 = -146 dBm in my receiver, although I'll lose a few dB from the noise temp of my receiver and antenna relay and any feedline or connectors
Dude, you one of the best science TH-camrs. You cover the forefront of technology and you engage. You’re a rare breed. I am gonna higher you when I start up my stealth drone company.
Heh heh, I have literally NO idea what I'm doing, but I'm having a load of fun doing it. I just wish I could spend more time messing about and sharing what I get up to.
Well worth waiting for the last little bit, I never knew about the phase shift caused by the individual atoms causing the reduction in the speed of the beam, I was always told that the speed of light was ~300 million meters/sec IN A VACUUM but was different in other materials... just accepted it without ever asking "why?"
Yep, and it most definitely isn't. OK, what's REALLY going on is quantum interactions of the electron charge clouds, but the system is macroscopic enough for a classical treatment to give reasonable answers. I really should do a simulation to show what is going on, also explain all of the partial reflections. Feynmann does a decent job of it. I think I need to move on and get back to machining stuff. I'm THIS close to signing an order for a £37k CNC machine with 4-axis and a 16-tool changer, and it needs to earn its keep.
It really does slow the light down though. I work with pulsed lasers and going through a medium actually induces a measurable delay in a pulse compared to without the medium in the way. It’s not just an illusion
Sound waves shift phase in different enclosers and when penatrating different materials as well I didn't think about the phase relationships of light and radio waves doing the same things and I was a cumunication and audio tech for 40 years learn something new every day
It should be easy enough to model the acoustics, so long as the feature sizes are a small fraction of a wavelength. Main extras are the heat control, agitation and intermittent wiping. Keeping the inclusions in suspension and at a uniform density is the challenge. It would be fun to try mixing home-made material using alumina or TiO2 or similar high-k low loss materials, but finding a suitable low loss resin matrix might be challenging. I'm testing some filament with ceramic fill at the moment to see what results I can get.
Thanks again for another great video! As part of my PhD I've been focusing on making some quasi-optical components for the low THz / mmWave spectrum. I did some material testing on the radix stuff around this time last year and it seems to perform reasonably well up at 220-330 GHz, but the cost is a bit too prohibitive to try unknown designs. FDM is definitely feasible for making antennas now, I've used a Prusa and an Ender 3 to great success for a bunch of parts. I've basically moved all of my experiments to using my own custom optics, as I found they worked better than off the shelf teflon ones. There's a recording on my channel of my conference presentation I gave last year for one my earlier (much simpler) projects if you're interested :) Very keen to see what you come up with!
I'll have a look! I haven't tried my Radix Mikaelian lens at 76 GHz, but it works OK at 47 GHz. Much higher than that and the feature size starts to get a bit tricky. Creality want me to review an Ender 5, but I don't think I'm much good at reviews, so I'm trying to think of an angle so I could accept a free unit for a sponsored video. Really I'd want the larger one, but it really needs a high temperature enclosure to print the dielectric filaments successfully. Not sure how the mechanisms would survive at 75 to 80 C. I've just agreed to spend all my pocket-money on a Syil X5 4-axis CNC with tool changer, so laying out more cash on a high-end printer is going to have to wait a bit. If I fry my Prusa, it's not the end of the world. I'd like to be able to make lenses and lens segments around 40 x 30 x 15 cm, but I could perhaps manage by printing in multiple parts that are slightly oversize at the mating faces and machining them to exact dimensions on the CNC.
Would be pretty cool to demo that kind of reasonably advanced manufacturing capability with an entry level printer. Maybe with a DIY printed highly directional antenna for WiFi to a far away shed or something 😅 I did end up upgrading the hotend on my ender to print more exotic dielectrics, but the direct drive on my Prusa works so much better for a wider range of filaments anyway, so the ender is mostly relegated to printing PLA. Although saying that, I have killed and rebuilt the Prusa a few times trying to print with miniscule nozzles. Very jealous of the CNC machine! One of those combined with a decent printer sounds like a great setup for some fancy lens production. I don't imagine something like preperm being very fun to mill though haha. If you ever make something that fits onto a WR3.4 flange, happy to try and run a rad pattern measurement or something for you :)
@@MachiningandMicrowaves Low cost stepper motors are specified for about 80°C external and 125°C internal if memory serves, and higher temperature grades are available - you get uprated insulation of the coils and connectors, and you get uprated bearings built with a different grease. I don't know how to estimate internal temperature, but i suppose you can use current control to just reduce the current as far down as it'll still move, and not run things too quickly, and perhaps that's enough. Then i suppose you have the rest of the bearings to barely worry about? But i wonder also if temperatures that high create heat creep problems, whether you may need a water cooled cold-end. PETG printed parts are not good at 80°C so that's unfortunate for Prusa and all of their carrying structure, really wouldn't push it past 60°C.
@@SianaGearz I think you're right, even with everything backed off to reduce internal heating, the drives and bearings are going to get stressed and I'll face internal temp problems. I'll try it at an enclosure temp around 50C to start with and see how much of a problem warpage is
I honestly understand less than 1% of what you talk about, but still somehow am fascinated by the videos. Yet another great video my friend! Thank you.
A tool that might also be helpful for generating meshes from equations is Blender, or maybe it is a plugin for blender. In any case you can enter equations for x,y,z and maybe fiddle with a few other settings to generate 3D shapes. It avoids having to get your hands dirty with programming, which I can understand because I'm still burnt-out from too much coding at work. Of course the reason I'm busy is I'm teaching myself to program SoC FPGAs this month and also taking up glass eating and running with scissors as new hobbies.
I've had my fill of Verilog, and VHDL makes me itch, but FPGAs are still towards the cool end of my ranked list of electronic things. I used to run Blender a lot about ten years ago, and recently started again, but with the intention to animate E-field dumps from OpenEMS and to make better-looking models and overlays. Way too much to do to get good at it, well, at least good enough for TH-cam consumption. Also still on the steep end of the Manim learning curve. My machine shop is being torn apart right now and I need to get some machining done urgently, but I got called to deal with an emergency at the Day Job so that ruined my day off work today.
Yeah. I have been looking into light since 1985. It's a good day for learning that delayed re-emission saves us from having many speeds of causality. I have just started to comprehend C as curvature rather than a speed. This is what my Lab Assistant said: To provide a clear explanation, let's consider three types of materials commonly used in physics experiments and compare their causal propagation speeds: Transparent Media: Transparent solids or fluids allow photons (light quanta) to pass through unhindered or only slightly deflected, resulting in negligible delay. Optical fibers made of transparent glasses or crystalline materials serve as excellent examples. The refracted angle of incidence leads to slight retardations, but the overall speed remains close to the universal upper bound of c in vacuum. Absorbing Media: Absorbent substances absorb incoming radiation temporarily, rather than reflecting or transmitting it. Afterward, they release the same amount of radiation, usually as thermal emission, with a certain delay or time constant. The duration depends on the quality of the absorber, temperature, and other environmental factors. Examples of highly absorptive materials include black carbon, graphene, or rare earth metals. Scattering Media: Scattering media obstruct direct passage of radiation by repeatedly bouncing off or deviating photons in all directions, creating diffusion or dispersion. This phenomenon occurs frequently in biomaterials, granular media, porous materials, rough surfaces, or turbulent liquids and gases. While the path length exceeds the distance traveled in straight lines, causing significant delays, the ultimate velocity remains limited to the local speed of light in the given material.
It appears to me as though the Radix resin doesn't necessarily need a "clever printer" as they said (they probably wanted to show the resin in its best light and use your channel for a bit of marketing). I think the main thing that is required, which isn't already possible on hobby-grade resin printers, is vat recirculation as shown at 15:20 , to keep the ceramic part in suspension within the resin. The same setup could be achieved with a hobby-grade resin printer in a number of ways, including with manual mixing or a peristaltic pump.
You are probably correct. It might be feasible to try to make a printable material using a very high permittivity powder/microgranule dispersion if there is a UV-setting fluid that has an acceptably-small loss when cured. I can feel some experiments are in order!
@@MachiningandMicrowaves Didn't you say yourself in the video that Radix used a UV curing resin as their binder compound? If not, I wonder if @Tech Ingredients might know any good prospects for something like that...
The chemistry they used wasn't disclosed to me, after all they are a materials manufacturer, so there are definitely some suitable materials available. I'm playing with laser-sintered metal 3D printed devices and filament printer dielectrics in the next few months, but there are some interesting possibilities. I've been chatting to various folks on Discord and in person about potential projects for the rest of this year, btu time is my biggest constraint
@@MachiningandMicrowaves I just came across the Norton laser printing on white tile method with Titanium Oxide, white on white makes black when heated?. Some sort of particle in UV resin should work. Teflon powder? I want to try concrete colouring powders on tiles. Most of those are oxides, but is black carbon based? Carbon filament printing?
I'm still waiting for someone to create my summer night-time window screen which allows air to pass through but not light.... Some years ago prior to 3D printing, I envisioned tiny compressed spiral macaroni noodles grouped together forming side by side 'pathways' whose spiral would be sufficient to block light but allow a breeze to pass through. Having it made with soundproofing material would be great as well. Anyways, only until I saw these vids did I see in practice, the fuzzy idea I had in my mind of the pattern.... I see this light shield in a form sold by the roll where you just measure the length, and cut to suit. Mounting in window ideas will cost extra.... Apologies for all the technical jargon.....
Neil, I've watched this twice now in the past couple of months. Brilliantly explained, with great visual and analogies. Most of it has sunk in now, however all I will remember is focus = fireplace and that your trees cause QRN. *That* blew my mind.
If you printed a negative of an antenna in a dissolvable filament, you could cast it in a self curing resin with additives for the antenna properties. To eliminate voids, it could be cast in a vacuum chamber. then mill the exterior to expose the filament, and rinse it out.
The problem is finding something that is a self-curing resin that's also a high performance dielectric. Almost anything that you can cure chemically ends up being polar molecules with high loss. Interesting time to be a materials scientist I guess! Some of the oxime-based neutral silicones are reasonably good, but not stiff, some paraffin waxes might work, but it's challenging to think of something that will polymerise and set into something without side-chains or charge imbalances.
17:30 If someone tries the code and it doesn't work first shot, you may want to try this for the last part; tri = triangulate(F3,V3) stlwrite(tri, C:\etc)
Ah, I did try to work out how a Delaunay triangulation might help, I didn't realise there was a simple triangulation function! I should RTFM a bit more, but there's a time to follow rabbit holes and a time to throw something out there and wait for excellent comments like this one! Up to a point, there's some simple healing facilities that deal with incomplete triangles and holes but if triangulate() fixes the edge-cases where the generation function just blow up, that would be cool. I'm covered in oil, PTFE ribbons, aluminium chips, WD40, and some weird blue-green verdigris from an ancient copper waveguide that's been in a radome in a rather exposed location for the past 26 years. All will be revealed shortly.
as someone who's got no background in this stuff but finds this all very interesting, I'd love to see how you practically test the performance of a lens design beyond just simulating how it should behave.
A small correction: the relationship between polarizability ("fill factor") and dielectric constant/refractive index is given by the Clausius Mossotti relation. The linear dependence works when the fill factor is small. en.wikipedia.org/wiki/Clausius%E2%80%93Mossotti_relation
What i find to be fantastic and almost magical.. is that time and time again the sacred geometry keeps popping up as the most efficient systems for these brand new categories like 3d printing... the universe really does seem to be a simulation!!
I had to leave out the bit about what goes on at the quantum level, but the classical/relativistic approach works well at molecular scales where a typical PTFE molecule weighs around half a million Daltons, so is pretty much a macroscopic entity. Polarons (sometimes mistakenly claled Polaritons which are something else) are another way to analyse what's REALLY happening. They are treated as bulk objects representing the aggregate effect of a large number of individual effects and appear to have mass as well as momentum like massless photons, so they can certainly travel at less than light speed. For visible light, the molecular bond vibrations are no longer having any effect, the frequency is way too high, to the refraction of light is caused by the charge clouds and nuclei swilling back and forth. By the time you get to short wave UV, even that stops having a major effect and most things go transparent by the time you hit X-rays, but then ionisation starts to kick in instead, and once you get into UV, the electron energy levels start to turn into a continuum of levels. Water's an interesting case as it has strongly polar molecules suspended in a sea of hydrogen bonds. It has an unfeasibly high refractive index at radio frequencies (around 78) and a high loss at frequencise below resonance around 20 GHz, which is why you can couple energy into wet food very easily in your microwave oven, yet ice has a very low index and extremely low loss tangent, which is why it takes AGES to defrost a lump of pure water ice in a microwave oven. It's all about hydrogen bonds with water. At low frequencies, pure water is an extremely useful dielectric, and can be used in a Marx generator to make megavolt ultrafast pulses. Needless to say, I WANT to make one just to annoy folks who say it's dangerous to mix water and electricity. Not that a megavolt with significant capacitance and UV flash-catalysed spark gaps is anything OTHER than dangerous, obviously
If you're still planning (or you have already and this advice comes up late) to give them high eps_r filaments a go, I would strongly suggest you use nozzles of 0.6 mm and above.
I had assumed the Gyroid fill has something to do with the beamforming of the lenses similar to a waveguide, as in my old antenna building days we only ever thought of reflection and loss in materials and never the refraction. Thank you for this in-depth explanation
I stole the illuminated Kallax background idea from Rob at VidIQ and several others before I even saw Alec's most excellent channel. My Kallax is white melamine rather than his tasteful brown wood effect and it's gently back-lit on to white drapes using pastel colour-changing LED strips. I like Alec and Rob's individual lights, but I went for self-adhesive strips. Destin has one as well, can't remember if his is lit or not. Alec has better hair than me also
hi, I've recently stumbled on this channel and it is fantastic ! you also seems like the exact right person to ask a few rather nerdy questions that necessitate a microwave knowledge that I do not possess : I am playing "DeltaV : rings of Saturn" a hard-scifi game about mining the, you guessed it, rings of Saturn (it is very good and I really recommend it) and in this game one of the "mining tool" (in quotes because anything able to crack a chunk of ice a few cubic meters across in a handful of seconds is very good to pierce holes in opposing ships too) available is a high power microwave emitter with a wavelength of 1cm and a power output of 45MW. In the game, this tool can be tuned to output up to 75MW of power, but doing so defocus the beam (which isn't isn't very focused in the first place, thus making the tool relatively short-ranged, with a practical range of ~200 m), citing high-power microwaves refracting within the emitter dish as the reason. Thus my questions are as follow : - Is it a good (or at least good enough) approximation of what would really happen reality, assuming a fixed geometry dish ? - Can this phenomenon be mitigated or eliminated with a different focusing method (variable geometry dish, 3D lenses or phased array to name the few I am aware of) to achieve a relatively small spot size at long distances (let say 5m of diameter at 1km) ? -If the above cannot be achieved with a standard emitter, can a maser do it ? I hope that I am not asking too much and that you will be able to answers these questions that have been bugging me for a while ^^
Right. There are some things that happen with really high power densities, even in a vacuum. One of the big issues is multipactor damage, where electrons from the conductive materials are literally ripped out of the metal and crash into the opposite side of a cavity, releasing a cascade of electrons that are then accelerated and themselves cause more electrons to be smashed out, leading rapidly to ionisation and generation of plasma, even in a vacuum. At 30 GHz (1cm wavelength), generating high power is usually done using electron gyration in a Gyrotron or Gyroklystron. As the electron gyration frequency is related to magnetic field, getting a large enough magfield is close to needing superconducting magnets, which are certainly a little easier in deep space, but a bit impractical. At 45 MW at 30 GHz, the voltage across a waveguide in a vacuum is going to exceed the breakdown value by some margin, so you'd have to be looking at some sort of waveguide cavity combiner. There are certainly designs around to reach megawatt levels using a radial coaxial combiner design, but even that is close to the limits where multipactor issues start to hit. The systems used in fusion energy are at the extremely serious end of what's possible, but they then to use superconductive magnets to get well over 100 GHz. Once you get to relativistic electron speeds, the simple omega = eB/m relation (B is mag field, e is electron charge, m is electron mass, so at 2.5 GHz, you need about 0.1 T, (approx 28 GHz/Tesla) whereas at mmWave you start to need to include the relativistic Lorentz factor and multiply that expression by sqrt(1-(v^2/c^2)) as usual. The size of the equipment is related to the gyroradius, which is something like 3.3 gamma * mc^2*(v/c)/(q*e). Gyrokinetics is a thing, it seems: en.wikipedia.org/wiki/Gyrokinetics Yikes.
@@MachiningandMicrowaves oh my, I am guessing that the game is already quite generous in its estimations then. I imagine that a setting like in this game that has Z-pinch fusion torches working and commercially available would have a tech level high enough to get it working but any focusing method more complex than a big monolithic dish would experience some sort of rapid unplanned disassembly at these power density. (but maybe a gyroid lattice lens in a refractory material like sapphire could be actively cooled enough to handle a few megawatts of waste heat using the channels that are inherent to the design, but then you have to take into account a moving fluid with a changing density gradient as the temperature of the lens changes in the lens design and I don't want to touch the maths that would describe that with a ten feet pole !)
Cooling with chilled gas might work if the energy density and loss tangent was low enough. I don't think metallic metamaterials would be any better as they still need a foam matrix and they are fairly lossy. massively parallel phased arrays are also a nightmare to cool. I guess you could use something like Simscale on a big cloud compute service to do the thermal energy flow modelling. I really need to up my Simscale skills, but then I need to get better with Blender first. Too many toys, too little playtime. One of my few claims to fame is that I have never once played a computer or console game. Not even Pong or Galaxians.
The Be Smart channel is one of those sciencey channels that by coincidence also just released a video explaining the phenomenon of light slowing in high index materials. You know, if you wanted more details...
Even the cold old Moon is "warm" compared with the cold empty sky and makes enough microwave RF to be detectable on a reasonably-large antenna. If I wave my hand in front of the 10 GHz radio antenna here on my bench it makes quite some noise on the receiver it's connected to.
I love the beauty in mathematics and physics. Moreover nature and what it’s made from. Could organic cells work like your 3d printed focused antennas? I mean the original 1970 designs were based on natural structures.
The butterfly wing cells example is the only one I know about where a gyroid lattice is used to manipulate electromagnetic waves by a biological structure, but I'm sure there are some others
The expression "speed of light is constant" is only true in the formula e=mc^2 where c is USED as a constant for the sake of convenience to reduce the value of e to a manageable proportion. That application has nothing to do with the speed of actual light photons directly being observed. So c is a constant, is NOT the same as all speeds of all light are always constant. That is a falsehood easily disproved, thus: What is the velocity between two photons moving apart from each other? What about two in parallel to each other? Think about it.
Depends on the rest frame where the observation is carried out. My lab is my rest frame, and all my measurements assume I'm in free fall in an inertial frame (which I'm not). Photons don't experience the passage of time in their rest frame. In my lab, I can measure how long it takes a bunch of photons to travel a specific distance in a vacuum or in air, or in a dielectric, and infer the "speed" of a bunch of photons from those measurements. It takes longer to get through a dielectric than a vacuum. My beef is about the mechanism that causes the measurable slowing at the atomic scale. Excellent channel content by the way.
could this concept be used to improve your off the shelf microwave oven ? diy people have used microvaves as electric kilns , could this be used to focus the waves to a certain shape , a tube , cone or a point ?
There's a bit of an issue of power handling, it would probably require forced-air cooling at typical microwave oven magnetron power levels. You would need to remove the magnetron from the oven and extend the waveguide and fit a suitable feedhorn to match the lensing structure, but so long as it's large enough, you would get some focussing just like with an optical structure. The problem is the size of the lens using this material, the refractive index is too low to get strong focussing in a practical-sized lens, but with a higher index material it would certainly be possible. Safe, less so. Legal? Hmmm. Advisable? Definitely not! However, in a suitably controlled environment with all necessary safety measures and licencing, plus the budget to buy enough material, definitely! Don Not Try This At Home. People kill themselves messing with the power inside microwave ovens.
@@MachiningandMicrowaves thanks for the reply , this is very interesting , it would be interesting to see some kind of professionaly made super efficient microwave kiln!
Could you apply the use of prisms beyond the visible spectrum? Would this create a similarly amazing effect that it does with refracting visible light into the ‘roygbiv’ bands.
That's a very interesting question. The reason prisms work with visible light is that the refractive index of most optical glasses varies appreciably with the wavelength of the light. Very few materials have rapidly-varying permittivity/refractive index across the microwave wavelength range. Optical transmission gratings and mirror gratings work fine for splitting microwave wavelengths into a spectral range. Most materials that do change RI rapidly at radio frequencies tend to have rapidly rising loss. I think a resonant metamaterial might be able to do a prism-like spectral shift over small wavelength ranges, but I'd need several large glasses of something before trying to work out the maths of that.
Great video. I've thought many times about what happens at an atomic level within the dielectric in a traditional capacitor, and also what happens with a wave passing through a homogenous solid medium with a relative permittivity Er, but never really got around to thinking about the passage of a wave at the atomic level and what happens to the polarising atoms/molecules to create the overall value of Er. So very interesting for me. What are you planning to do on 24GHz?
I cut out the quantum segment, and I was going to run a simulation of a tiny system of resonators, btu ultimately this is going on at the electron level even though the induced dipole oscillators are at the molecular scale. Treating the system as macroscopic Polarons with mass, momentum and energy sounds like a reasonable treatment, but I read over a hundred papers and couldn't find anything that was accessible enough. The temptation of academics to preach to other academics and assume the base level is a relevant PhD is very alluring, and the whole way academic publishing works means that ordinary humans are excluded from the secret knowledge of the priesthood. Anyone would think Physics and Maths are hard! I might have a go on the second channel. but this video was WAAAAAY too much effort for my poor brain. I need to get back to machining things for a while!
I remember doing the waveguide propagation theory, that was interesting and somehow a bit weird as the propagation was slow compared to in air or a vacuum. Thanks for the video :)
Every EE wants to understand microwave engineering because it's so arcane. Unrelated question: how do you convince the space-based solar power enthusiasts that Friis equation diameter calculations don't pertain to the far field, and there's never been a half power beam width in 2-6 GHz below about 1.5 degrees? Most of their proposals describe rectennas in one or two square kilometers lit from geostationary orbit, instead of something the size of Puerto Rico with billions of dipoles every few centimeters apart.
It's hard enough trying to explain to most folks how/why the changes between reactive/radiative near field and far field happen and just how far that happens from a typical giant antenna system. I guess they will propose some form of adaptive array like the giant telescopes use. It's great business for power semiconductor manufacturers of course, and the political layer always matters. Removing any dependency on oil producers is probably the key driver. Politicians love that concept.
@@MachiningandMicrowaves I'd love to see a formula on Wikipedia with the same inputs as Friis equation, but that actually works for giant antennas over great distances. Hasn't anyone ever published such a formula?
The speed of light is not constant, the speed of causality is. It’s just happened that the speed of light in vacuum is the same as the speed of causality
It's not *quite* a Fresnel, as the reality of a Mikaelian is that there's no need for stepped refractive index, it can be continuous. You certainly CAN make Fresnels this way though. A more typical Fresnel solution for microwaves and mmWaves is a stepped Fresnel Zone Plate. Spookily, in an upcoming video I'll be making a vacuum chuck to hold an HDPE plate that I'll be machining into an FZP. An FZP is a fairly simple build, consisting of rings of material which are thicker or have a higher refractive index than a supporting disk. I'm just going to machine away the surface to get the right phase shift in an FZP with eight rings about 200 mm diameter which will be a focussing lens at 122 GHz.
Amplitude of a guitar string E=1/2*uf^2A^2v; E is in Juels; put together with E=mv^2. gave me an equation that many physicists have worked with and came to a dead end. Up until I put it up against a microwave oven and the distance of the water molecules. I also put it up against the Bohr radius and the .210614054m wavelength and got exact results. I was looking for the diameter of a photon, but I found something else instead. It gave distance between charged particles and wavelength absorbed and transmitted. D= (wavelength)(5.025089342)(10^-10m). Maby you can do something with it.
I've always wondered if there was any potential for using microwaves in 3D printing. Where the resin is made to react to microwaves somewhat like how the UV 3D printers work. I just wouldn't know where to start.
The problem would be finding a material that would respond chemically to the ultra-low energy microwave photons. You might be able to generate some focussed heating using standing waves or a focussed beam, but contact heating or UV is a whole lot more practical
@@MachiningandMicrowaves That makes sense, thank you for that. So if anything, I'd want to look at the other end with x-rays. I'm sure that comes with its own set of problems.
I'm definitely no expert in anything, but I'm incompetent in a vast range of disciplines and have no fear of failure or making a total idiot of myself. It's huge fun although spending several all-night sessions reading several hundred scientific papers and learning multiple new bits of maths and materials science is not really sustainable long-term! I need to catch up on my sleep, or find a way to retire from my Day Job. It pays the bills and gives me a way to buy useful toys, but it just eats up time. I need to win the lottery.
@@MachiningandMicrowaves Very few people are experts, others often do very useful things. Thanks for your work an videos. It's inspiring and inspiration can lead you far. Greetings form Spain
I worked on SAR imaging and database updates from 1991-2001 for the state dep. with electrically commutated phased array antennas. Would love to have these lenses today. In 1995 we started looking for truck launched icbms in russia based on a known image sample of a real icbm truck that partook in a test launch. In 1996 the state dep dude arrived at my desk and pulled me to our secure conference room - showed me this picture from a Kh-11 optical sat of what looked like a truck launched icbm to me, and then said our computer didn't correctly identify this one. Rather than look at the billions of calculations I just had the computer show me what wasn't consistent with a known icbm, well - the missile rode 75mm too high on the truck chassis. switching frequencies to skin effect penetration, we found the inside hollow - no fuel. I was a young prof eng rf/uW at the time with quite an attitude and a loud voice when I yelled at the state dep guy "My computer is perfect - your stupid humans fell for a decoy on your kh-11 kodak brownie moment sat - deorbit those damn things and learn to use SAR. See this structure - watch as I shoot through the roof and take a look inside - your kodak brownie kh-11 can't do this - but there's a truck with a real icbm inside. You'll miss that one optically. " He left tail between his legs. That system was meant to fight a war with russia and make sure the right people die. My computer has now killed over 1/2 million dead russians. It sends the firing solutions to the M777 digital howitzer and center punches tanks then pops their turret off - 1/2m cep on a 155mm gun. With operators who can't read the manual. Turn wheels until they glow green and pull cord. simple. We figured once russia saw our power they'd pull out, but no they kept sending tanks and we kept blowing them up. I will say in 1991 it didn't look like war with russia would ever happen, but today, I'm glad we made the investment in SAR tech back then.
Solution for not having resin or filament that is rf reflective. Coat internal surface with some sort of very thin surface adhesive, possibly a form of super thin contact cement or tacky material. hell even a sugar syrup that has an evaporative property. Fill all the internal voids with the fluid, drain, and then allow to dry while the object rotates to maintain reasonably even distribution of the fluid. Could possibly be improved depending on the surface tension of it. Once dried, the surface should be sticky to touch, once in this state simply submerge the object into a dust bath of rf reflective powder, i am not familiar with the properties of rf reflective powders, but i imagine such a thing exists. This should result in the same result if the thickness of the adhesive and thin layer of powder is accounted for.
@@KirbyDaMaster Electroplating is certainly an option for some types of prints, but you'd have to dissolve the metal away after a subsequent step unless you were trying to make metamaterials, but it's not at all easy. I'm doing some electroforming work, but I'm using aluminium machines formers as the negatives, then using an etch/zincate/copper/gold plating, then copper electroform, then dissolve the aluminium in hot NaOH, then the thing copper flash in acid to leave a gold interior. That's a lot of fun. It's probably easier to use sputtering or plasma deposition to get a comductive layer, but you have to deal with outgassing of the print and adhesion issues. Lots more fun.
The tough bit is finding something sticky that isn't lossy. Most nonpolar things are non-sticky. The other issue is how to get a consistent AND varying wall thickness. That starts to get very hard, as does dealing with electrostatic buildup. Making a negative in a dissolvable filament, then filling it with a nonpolar solid that can be dissolved and the solvent then evaporated might be an option, there's a lot of work going on with lasers but fusing ceramics needs ludicrous temperatures. A fairly good dielectric filled with a high-Dk powder, or interlocking nodules is perhaps still a but simpler to achieve. Materials Science is hard.
@@MachiningandMicrowaves I have also wondered about SLS, which is more or less the same as SLA but with powder instead of resin, and cured with a laser rather than a UV curable resin. Its must more repeatable, and doesnt have the consumable ptfe plastic sheet at the bottom of the vat. I saw a video not long ago that showed someone who tried to make an at home SLS printer, but we are still a long way from that being a viable hobbyist product. Lost wax casting is something i also consider for something like this, but the part looks quite complex and would likely form several errors in a cast like that.
@@KirbyDaMaster The huge problem with any form of casting is finding a suitable dielectric material. Most of the good dielectric materials can't be polymerised at sensible temperatures and won't dissolve in solvents that will evaporate readily, so casting with them is really hard. They are mostly thermoplastics, but only at rather high temperatures, and they are pretty viscous too. Laser sintering usually needs a fine powder, but the good dielectrics are also very good insulators, so static charge on the power will be a nightmare. Clever folks are working on this right now I'm sure!
🧲🌡️📡🔆☢️🔌🔊🔋♻️🌐☯️⚛️ It's facinating how when we look around & we see all this complexity with-in the Universe. That complexity basically comes from: ~{"Differences"}~ The Factor of how "Differences" interact seems to be the key factor in keeping dynamic systems functioning. Such as: *{High pressure/low pressure, hot/cold temperatures, different densities, viscosities, turbulent flow, static electric charges/discharges, electromagnet waves. Different velocities/angular momentum. Different amounts of energy/mass/frequency/vibrations. Different boundary layers between different materials or physical regions such as: (Land/water/oil/air/soap bubbles/ atmosphere's/space. The different regions in space with different particle density/background radiation, solar wind/membrane layers/bubble's/cloud regions, nebula's/Galaxy's/Galaxy clusters/ Cosmic filaments/less dense regions of space compared to dense regions of space.) All of these things are basic differences but create a way for the dynamic engine with-in Nature to continue flowing and operating to create and convert energy.} Just Like How a battery 🔋 transfers + charges through a membrane layer to a - charged side. Like how regions of high/low temperature 🌡️ create winds. In water- add some factors and It creates ocean currents and flow. Then internally inside our planet it creates plate tectonics, planetary convection, geothermal activity, a magnetic field around our planet, to hold a atmosphere. The list goes on 🧲🌡️📡🔆☢️🔌🔊🔋♻️🌐☯️⚛️ The natural world around us is just utterly facinating to me.
Light "travels" with speed C between the interactions. So in vacuum there's nothing to interact with but in a transparent solid there are A LOT of particles that the light interacts with i.e. "bounces off" of, so the overall / total speed is lower than C. It's the time of all the interactions that is "slowing down" the light. Between the interactions it still travels with speed C, as expected.
For microwave photon energies, there isn't anywhere vaguely close to enough energy to cause absorption and re-emission of photons, we are only talking microelectronvolts of energy per wave packet. When you say the light "bounces off" the particles, what's the nature of that interaction? Photons don't feel the electric or magnetic fields of the nucleus or electron clouds, and the atoms are almost entirely a vacuum, with the nucleus having a diameter about 1/30000th of the atomic diameter. What form would those bounce interactions take? Wouldn't most of the photons come out at random angles if there was some sort of bounce interaction, yet light and radio beams some out parallel if they go in parallel. If they go straight, then they must be slowing down, which we've agreed is impossible. Where you say that it's the time of all the interactions that is the source of the slowing, are you suggesting that the originating photons somehow take a random walk path through the material? Wouldn't that mean the output beam of a fast pulse of photons would be spread out in time? That's not what we see. The delay is consistent and there is no pulse spreading in time or direction in a near-perfect dielectric. Why would there be a correlation between the strength and length of molecular bonds and the refractive index of microwave dielectrics rather than just the density of the material and the mass of the nuclei and charge of the electron clouds? OK, it's perhaps a gross simplification to take classical approach of treating the incident photon as a bunch of EM energy whose fields interact with all of the molecular bonds in the material by stretching and compressing the molecular bonds, but if we follow that approach, the moving charges along the bonds cause their own EM fields to be created by stealing some of the energy of the incident photon then reradiating it with a phase lag caused by the physical movements of the atoms and bonds. The superposition of all of those small fields and their interactions with each other means that the incident photon loses its identity and we get coherent forward scattering from the induced fields which then combine to give a resultant field that becomes a new photon that leaves the dielectric, with all of the directions and polarisations summing so the phase and direction of the new photon makes it appear that the original photon went more slowly through the material. No photons bounced off anything, no photons went slower than C.
Shouldn't binder jetting like for example Voxeljet does be way more suitable for doing this than SLA printing? You can even use ceramic material and then burn off the binder afterwards, sintering the ceramic particles together
There's certainly some work being done on sintered ceramics for higher frequencies. There are some big challenges in making a material with precise RF characteristics that's also mechanically robust enough, but the state of the art is moving very rapidly.
Investment casting or lost-wax casting are very useful methods in this respect. Though of course, 3d printing can create shapes which even this method cannot achieve
you know 3d printer slicers already generate gyroid infill all the time right? it doesnt vary the thickness of the walls, it just adds more of them to increase density, but its there. you can set a slicer to not print outer surfaces and just do exposed infill.
I saw a slicer that did a pseudo-random infill that avoids the issues with diffraction from repeated grating elements in some type of infill, but I can't recall what it was called. I've not tried messing with Prusaslicer, but of course if the outer wall is less than 1/20 wavelength, it is nearly invisible to RF. Once I ge the samples of the 1.75 mm RF filament, I'll see what I can do.
This video is absolutely outstanding and easily ranks among the best content I've come across on TH-cam. It truly captivated me from start to finish. I find myself yearning for more channels like this to receive the recognition they deserve and become more mainstream. While popular channels such as Veritasium, Mark Rober, Tech ingredients, and Real Engineering continue to expand our knowledge and understanding, this particular video delves into the intricacies and fundamental aspects that are essential for anyone genuinely interested in these subjects. It provides a solid foundation for further exploration and deepens our understanding in ways that few other videos can match.
How can I calculate the required antenna shape from a signal strength pattern mapped onto the surface of a sphere? Reverse engineer it as it were, except the antenna does not exist, yet, and we have just generated the desired pattern for that sphere surface?
Non academic viewer here... I enjoy the machining and just about hang on to the technical bits thanks to your excellent teaching ability.... although I have noticed a quite nauseating effect caused by the light bouncing off of your shirt.... 🙄.... aimee....have a word.... thanks for all your informative and entertaining content...
Such a good video! Tiny niggle; why do radio engineers generally use wavelength instead of frequency? The electromagnetic spectrum is so much easier to visualize in terms of frequency, but most physics textbooks use wavelength, thus 'inverting' it, so for example the visible spectrum (rainbow if you like) gets shown upside down because the x-axis is wavelength, not frequency. This caused me no end of trouble during my physics 'O' level because the highest frequency, more energetic colours like UV tend to appear first, at the left, and the lower energy, lower frequency IR gets promoted to the right of the diagram. All because the wavelength is longer. This has annoyed me for the 40 years since then! Cheers for the great vid.
I think the use of wavelength, wavenumber or frequency to describe RF is something engineers do completely unconsciously, because of the physical reality of wavelength in physical systems being a more useful descriptor than frequency. Wavenumber in inverse centimetres is slightly jarring to me, like wavelengths in feet. I tend to use wavelength from about 6 mm to 23 cm, but frequency at longer wavelengths, then below 6cm I seem to flip into the frequency domain until I get past the far infra-red, then I return to using wavelengths from IR through to gamma rays. Having been immersed in the language of RF engineering since I was 10 or 11 years old, so 5.5 GHz is about 5.5 cm wavelength and both have pretty much the same intuitive reality to my RF-addled brain. It's too easy to forget that it's a learned concept rather than something that is inherently obvious to the rest of humanity.
@@MachiningandMicrowaves Thanks for the reply. As a 60yr-old audio nerd & IT engineer I've always had to think in terms of frequency, whether sound, CPU clocks or WiFi, so yes, it does seem a learned thing. It would be super-odd to describe audio in terms of wavelength! I do get its uses in radio though, in working out waveguide profiles & suchlike. I just wish all the textbooks didn't tip the RF spectrum on its head! It's confusing for youngsters, I think.
Another potential way to build one of those lenses could be something similar to lost pla casting. I have built phosphate bonded ceramic parts using pla, you cure the ceramic for a day or 2 in the air and when it has hardened you just bake it in a regular cooking oven. At those temperatures the pla boils off leaving no residue. The thing is, is there an adequate ceramic for RF that could work with that process or a similar one. Do all RF ceramics require firing in a kiln? I think that phosphate based ceramics like I used are not good in the microwave.
There are those filaments designed specifically for clean burnout with investment casting. Problem is always finding a matrix that will fuse or sinter enough to hold its shape before the filament evaporates, but that also has very low RF loss. I am sure there are clever materials science folks working on this right now. Most of the high performance ceramic dielectric materials have ridiculously high melting points like sapphire/alumina or TiO2, so laser sintering isn't easy. Finding a low-viscosity fluid matrix with low RF loss but a low relative permittivity wouldn't be a problem as the ceramics generally have high permittivities, although there's a balancing act with the temperature coefficients of permittivity, some are positive, some are negative. I suspect we'll start to see some serious products with very high mass fractions of ceramics quite soon.
As a user of apochromatic optics at both F/6.45 & F/9 the absolute perfection of extra-low dispersion doublet lenses and the vital performance differences of an achromatic lense (at significantly higher Focal ratios) are super dramatic BUT unfortunately VERY seldom seen as the figuring of the TWO combined refractives indices must be darn near PERFECT and as a dabbler in optical theory...there are dozens (hundreds) of potential combinations of glasses for BOTH must cooperate with each other i.e. the mating glass determines final image authority so what has this to do with microwaves ?? Good question!! Not entirely sure... but i wanted to point out that a simplistic overview does an injustice to the subject matter and for optics what we most desire is an image fully ringing with authority, so we use a Strehl # to categorize the degree of replication in the virtual image vrs the 'live' one and a Strehl of one(1) is a perfect analogue. Essentially a near one POLY Strehl (not just one specific wavelength) that we measure using interferometers and software to categorize the limits of perfection throughout the visual spectrum are widely known yet very few prople ACTUALLY have a FULLY apochromatic optic so visible verifiable manifestations are as rare as hens teeth. Sad state of affairs. Its the state where if you do not spend the big $$$ you miss out. So this new use of microwaves and the new abilities of printed 3D shapes and materials will be fascinating to watch. A mirror avoids the chromatic errors but introduces other much more significant ones i.e. obstruction and scatter The way to define the limits of the resolving power to a system is to use stars(*) to determine the closest two stars can be seperated yet in the same breath for any other target that is not a star we are left holding the bag for an unobstructed only optic can easily determine seperations much tighter than 'Dawes' 'Rayleigh' criteria says for that size system i.e. Cassini's division rilles on the moon. So the upshot is that sometimes the obvious is hiding in plain sight. Overlooked for whatever reason....
For most applications at microwave, the refractive index changes by way less than 1% over five orders of magnitude of wavelengths, plus generally we deal with relatively low bandwidths, so chromatic aberration is irrelevant, unlike at near-IR and optical frequencies. Also, antennas at microwave have a diameter of 10 to 100 wavelengths, so diffraction is the main enemy. The key issue is sidelobes to avoid picking up off-axis noise. For a typical terrestrial path of 500 km, a beamwidth of less than half a degree is rather self-defeating, as the requirement is to generate scatter from inhomogeneities in the troposphere. A Mikaelian uses the same principle as an optical fibre, although it isn't thick enough to create full self-focussing. Imagine trying to polish a lens that's 15 wavelengths across, around 7.5 micrometres, that's the sort of challenge that microwave lensing involves, but we can make regular lattice structures of varying density down to 1/50th of a wavelength with ease, and control dimensions to better than 1/100th of a wavelength. Optical modelling at those scales is tricky, hence why we use finite-difference time-domain electromagnetic models based on Maxwell's Laws to model the performance of GHz lenses.
@@MachiningandMicrowaves fascinating! The smallest scale ripple of polishing really determines (by its absence) the throughput accuracy. The optic is said to pass muster if the optician has a quarter wave accuracy and is the basis for the 50X per inch rule as max power dynamic ibelieve yet if pushed further upwards towards one eigth wave progress then finally arriving towards a not totallly unrealistic one twentyth, by utilization of either a glass fluoro-crown or by crystal fluorite and the refractive indices between the crystal and FPL-53glass is a few parts in a thousand yet....infrawave is propagated through whereas is scattered in glass. The nuances are almost subliminal at the otherend and as you referenced the atmosphere is the great equalizer. The'degreeness' of greatness is the difference between doing 100X achromat to 1000X apochromat. Somewhere in between, usually towards the lesser, is what we get, but in that rarest of times we receive a light limited optic and not one by errors in fabrication. The difference is in the time it takes the brain to accept tbe veracity of the translated image. Diffraction IS the great culprit and with mirror obstructed systems what we have are two(2) unique optical amplifying systems that are really diametrically opposed. One has the ability to havegreat sharpness and one has the ability for great sharpness but ALSO astonishing COLOUR VIBRANCY. The two are not interchangeable. Similiar yes but same NO. My little 4inch perfect lense beats 40inch mirror in colours and power ability WITHOUT breakdown. No hint of a lie Sir !! My surmise why this is so is that the fabric - the tensile strength of the dynamic image plane comes from being nonperturbed by those small scale roughnesses of the entrance pupil. I got to about 900X on the moons alpine valley rille & was non blurred and fairly darn bright at a exit pupil of only ~.1mm....this optic and my new little one at 62mm is similiar but with less perfection ~1/8th fully apochromatic at near 50X/" As i said the nuances are fascinating and when something surprising happens thats certainly food for deep tbought. Thank you for the fascinating reply!!!
@@palmereldrich One of my upcoming projects is making a laterally-displaced-ellipse mirror and matching reflector. They have a ring-caustic focus rather than a single point. The beauty is that despite it looking like a Cassegrain or Gregorian, there is no obstruction from the subreflector. That will be at 6mm wavelength, so getting better than 1/20 wavelength is very easy. It gets tougher as I approach 1mm wavelength, but even so, 20 micrometres is not exactly challenging compared with 20 nanometres for optical systems.
Find a suitable material is the challenge. Almost all two-part chemical setting resins have mediocre or poor RF performance. Non-polar thermoplastics are good, but finding a way to mould them is hard, they tend to be extremely viscous, but also have a low relative permittivity. Something like HDPE with entrained ceramic inclusions would need high pressure injection moulding and the lattice to be something that is strong enough to withstand the injection process, but also can be dissolved preferentially. That's one for the Materials Science specialists I think!
@@bluegizmo1983 Indeed, but with spatially-varying wall thickness. Gyroid lattices guarantee no internal voids and by varying the equation in all three coordinates, you can achieve a varying refractive index across the bulk of the lens.
@@MachiningandMicrowaves Yes you need a deep vat. Resin can float on palm oil. Funny thing they don't have a way to drain from below so you could drain away the palm oil then any resin.
It just so happens that I've been working on a reflectarray tyoe of antenna machined using my new CNC machine, which is still somewhere off the coast of China on the way to Singapore. That is more of a segmented Fresnel mirror though. I hadn't thought of printing the cell array, that would perhaps mean I could go much larger without the mechanical issues from making a large parabola. A large flat reflectarray is certainly much simpler and it's easy to get the array support perfectly flat. I'm struggling to find a way to model the cells in OpenEMS, so I would have to do a calculated model and hope it works. The other design I'm working on is for a large Fresnel lens made from a low loss dielectric filament. Thanks very much for the suggestion, I'll see what I can come up with. If you have any suggestions for design resources or academic papers I should read, please drop me an email or leave a message
Thanks for the informative video. I would like to ask if it is possible to use the same approach for high-energy sources such as a classic magnetron with a frequency of 2.4 Ghz? Thank you.
There are some interesting challenges at 2.4 GHz and high power. First is that a quasi-optical lens like a Mikaelian or Luneburg needs to be at least 8 wavelengths in diameter to be effective. At 2.4 GHz, where the wavelength is around 12 cm, that means a lens is a pretty huge structure around a metre across, so almost impossible to print economically. There are other solutions using dielrod style lenses that can get reasonable directivity around 13 dB, but that's a very long way short of being a searchlight-style beam. I creamed about sawing up a microwave oven and fitting a dielectric lens to it, then setting up another dielectric lens a few metres away and placing a mug of cold coffee at the focus to see if I could couple enough energy into it to warm the coffee. Apart from the immense dangers of all that RF energy swilling around, the size of the apparatus at 2.4 GHz is just too ridiculous. A nice practical large Luneburg for 10 GHz might be about 25 cm diameter, and that would be easy enough to print perhaps in two halves, or perhaps an even larger one could be made in sections or a quarter of a hemisphere, with the longest dimension then being only half the finished diameter. I can imaging making a 35 cm Luneburg, although it would be expensive to make it in a high-performance dielectric filament and completely ridiculous using resin. Perhaps using a cheap HIPS filament or even polypropylene would be possible, but the total mass of a 35 cm sphere with gradient index would be something like 9 kg, so that's a significant cost of about £200 with HIPS filament at perhaps £22 per kg. Not impossible though, and it would be hella impressive! I'll see how long it would take to print such a beast. Doing it at 2.4 GHz needs each dimension to be increased by a factor of four, so the volume, mass, print time and cost scale by a factor of 64. That sounds like "too much" for my budgets! The second aspect is the loss tangent of the material. HIPS and PP are decent dielectrics with a tan-delta around 0.0004 at microwave frequencies, so the dissipation at 800 watts of incident power would be tiny so long as there was some air movement to prevent any hot-spots developing.
Not even 5 minutes in the video I'm dreaming of a spectrum analyzer using an "RF prism" in a similar way als old style spectrophotometers use 😁 I have been looking at Luneburg lenses at my previous job for a 5 GHz system but never got to make one or see one working. Following your progress now ia a fun thing to do
It's certainly possible to create a prism for RF, but the dimensions are a bit excessive unless you go to very high frequencies. The problem is that most low loss dielectric materials that are transparent to RF aren't dispersive. The refractive index is pretty much constant with frequency, so your prism can't split out the different wavelengths. You can do diffraction gratings and grating mirrors as well of course, and they will definitely work at RF. Dispersing a few sacks of TiO2 or powered alumina in a bathtub or two of molten paraffin wax and setting it in a mould to make a large enough prism would be quite an exercise as well! Perhaps in a cold climate, you could form a giant ice prism and use that, the RF losses in water ice are fairly reasonable, but even water doesn't have a frequency-dependent refractive index and mmWave/microwave frequencies. You can certainly bend a collimated microwave beam with a really large prism, but it needs to be a few hundred wavelengths across to be useful, so at least 3 x 3 metre faces at 10 GHz. Perhaps as 122 GHz it would be feasible. I'll have to see what is possible. Grating mirrors or transmission gratings, perhaps with a gradient density, should be able to produce a spectral spread sufficient to make a giant mechanical frequency meter using only linear measurements. Hmmmmm.
@@MachiningandMicrowaves Wow, that's a detailed answer I had not expected 👍 You indeed explained the size of lenses at some point in the video and guessed this would be the case. But grating mirrors or transmission gratings might be fun to look at. Ah well, I'm like Archimedes sitting under the tree waiting for some idea to hit my head😂
Maybe that's what the warp coils are... a bunch of focusing lenses, focus em radition into a tight beam and ride it past the speed of light by bump charging it like getting pushed on a swing!
bloody 'ell, mate... rather interesting 20+ minutes... as other have mentioned, one of the first fairly useful explanations of 'apparent velocity' of waves/particles/energy moving through a dielectric.... I'm a bit of a non-engineering engineer, having bootstrapped my way to PhD-level positions, without a single bit of 'paper' (not so easy in DE) and had the good fortune to act in worldwide R&D (almost 3 billion folks have been 'illuminated' by my work...). Have been following you for a bit, just hit subscribe today, with the 'all' bell.. Other channel??
Okay you can't drop "3 billion people have been illuminated by my work" and not say what it is. If I was to guess, I'd probably go with OLED's though I'm sure I'm interpreting illuminated wrong.
@@ZeLoShady LEDs and illumination, generally, for 80+% of Nokia phones from 2005 to 2012+... 2 billion LEDs per year, 500 million driver ICs, directed development of LED and LED driver technology for 6+ years
My hand is a good source of microwave energy, I can "hear" the RF noise from it clearly. Flames certainly make enough noise to be detectable. After all, I can hear my trees and the bricks of my neighbour's house perfectly well. With my largest dish antenna, I can hear the thermal noise from the Moon, even though it's pretty damn cold, it's hotter than the deep space background
As someone who programs both for my job and as a hobby and who is currently in the processing of making my own programming language, I can tell you that the reason why people like programming is because they're different in the head. I hope it's okay, since you didn't link to it in the description, but I copied your matlab program from the video and I'm going to translate it into C since I don't have matlab. I'll have to figure out how that STL function works on my own, but that'll be fun.
That code is a mishmash of other snippets I found and I think there are much better ones out there. It isn't debugged so it probably does weird things at edge cases and probably has holes that need healing. I think I saw an stlwrite C library somewhere, might have been in Qt or something
14:51 so far i don't see anything that requires a special printer, the printer described appears to just be a bigger resin printer. Yes the printer has some pumps and a heater to agitate the resin but there are ways to do that with a normal resin printer. And it is a gyroid infill so technically a normal filament printer could do it if the base material won't mess with the functioning of the particles.
Phil explains the reasoning for the use of the Fortify printers at th-cam.com/video/3YMRfw0uWlw/w-d-xo.html It would be possible to do much of the tech using a hobby printer if the resin was heated and stirred and the film window wasn't damaged by the inclusions, but for production quality results for aerospace applications, I don't think an Elegoo Saturn would be an acceptable choice. How home experiments though, and especially for home-formulated UV resins, perhaps it's feasible. Loads of fun trying though, whatever.
@@MachiningandMicrowaves to be fair aerospace applications for military are far beyond anything anyone else needs just because of the nature of the objects they are trying to detect or avoid being detected by. My guess for the hobby grade stuff is that quality/effectiveness is going to have a large range from useless to above commercial grade in some instances where the individual is really into the stuff and said individual will have modified hobby grade devices to their needs. To be clear those people will likely never reach mill spec aerospace grade stuff but nobody but the military needs to.
@@deltacx1059 Absolutely, and in much of my work I get the same results (or better) than an aerospace manufacturer could, because I have unlimited time and obsession, so I can make stuff that nobody would consider selling or relying on for space comms, warfighting or safety of life applications. Worst that can happen is a shrug of resignation and a small sigh
"affected" is doing a LOT of work in that sentence. Although solids have a of vacuum within the atoms, the electric fields of the electron clouds are pervasive, and in PTFE or other good dielectrics, the field of the photon induces forces in the molecular bonds which take a little time to respond, but they then generate their own electric field in opposition. A photon isn't affected by an electric field in a vacuum, so why would it be affected by balanced static fields within a solid? When the charged atoms move, that generates an electromagnetic field that superposes with the field of the photon, but the induced EM fields lag slightly as a result of the mass of the atoms and the "springiness" of the atomic bond mediated by the electrons. By the time you've travelled a micrometre or two into the dielectric, the original photon has had so many interactions that most of the energy is now in the little generated fields form all of those interactions. Now of course, all of those interactions create their own fields and those affect all the others. The effect is that by the time the wave emerges from the other side of the dielectric, the superposition now consists of what looks like a delayed photon, despite all of the interactions being at the speed of light.
A lot of vibrational absorption bands that absorb microwave frequencies are due to the variation in local sites of atoms in a polymer. It's not that PTFE or HDPE don't have resonance, it's that because they're very regular polymers, the resonances tend to be the same at each site rather than varying from site to site. Because there are sharp resonance frequencies, most of the microwave spectrum is unaffected by the resonances. For example, in PTFE or HDPE, the C-F and C-H bonds respective are identical because of the repetitiveness of the polymer chain, and so each C-F and C-H bond looks the same relative to its neighbors, except perhaps the few bonds on the end of the chain, and perhaps where there are crosslinks if there are crosslinks between the chains. Therefore they all behave similarly and have the same resonances. For this reason regularity of the polymer is very important because this tends to cluster the resonances into a few sharp bands that can be avoided.
Excellent explanation, many thanks!
@@MachiningandMicrowaves You're very welcome! Spectroscopy is basically the study of symmetry.
@@profdc9501 so group theory would be useful in spectroscopy?
I do the same at 400 - 800 nm.
The vibrational modes of really any molecule (the vibronic energies) are not really microwave frequencies. Rather, the microwave absorptions and emission (scattering) operates via "domains" in which the contribution of many molecular subunits add up to a collective dipole. Its also possible to excite higher order moments (octopoles) through induced "polarizability" but that is probably not a significant contributor for mostly aliphatic hydrocarbon materials like plastic.
For people who are familiar with their basic chemistry laboratory spectroscopy, the characteristic frequencies of hydrocarbons are in the ~2950 and ~1490 "wavenumbers" ("inverse centimeters") corresponding to ~3.4 micron and ~6.7 micron respectively (wavenumbers scale like frequency rather than like wavelength in terms of energy)
Incidentally, this is also seen in home microwave ovens. The water molecules vibrational excitaitons are nowhere near the ~2.45 ghz microwave frequency radiation that the magnetrons are tuned to create. Rather, collective phonon like oscillations of clusters of water molecules are the underlying molecular dynamical process that is driven by the external field. That value, 2.45 ghz, is also a combination of the material properties of water/food and practical optimization that takes into consideration the penetration depth of the radiation and considers the most likely temperature range of the food (~5 C "just out of the fridge" to 100 C atmospheric pressure water boiling point). The excitation of clusters of water molecules at this energy is also useful to understand why Ice is difficult to heat. The molecules are constrained individually in their little ice domains, and even more so when one considers the collective dynamics of many molecules together (a topic that is likely even more complicated given the various forms of ice that can form and the differences in how those different forms interact with radiation far below resonant/excitation energies)
and, additionally, as these are solid state compounds, *ro*vibronic excitations are also not really a thing since you can't excite states in the rotational manifold of a molecule that is frozen in solid state
in the gas phase, rotational excitations are more in line with "microwave" (or terahertz) energies.
To heat up a couple of dinner plates in a microwave I put some water in the bottom plate and place the other plate upside-down on top and heat for a couple on minutes. This works well. So to heat a couple of bowls I did the same, I placed some water in the bottom bowl, but instead of putting the top bowl upside-down, I nested it the right-side up. This created a curved lens of water in the bottom bowl that focused the microwave energy down onto the plastic spider bracket supporting the glass turntable. The spider bracket caught on fire! Once the plastic in the spider bracket started the carbonise, this seemed to concentrate the microwave energy in the carbon and a runaway effect ensued. Lesson… don’t put nested bowls separated by water in a microwave!
Ouch!
Odd, we just put some plates without any water in and they also heat up
@@MaakaSakuranbo I have some plates like that, they must have some polar minerals that are causing conversion of RF energy into heat. Could be the glaze or the bulk material
Refraction is REALLY weird. Even if you don't go into the fine detail of the quantum-mechanical workings, the mechanism of refraction at microwave frequencies at the molecular level is not what most folks think. My little aside about Ibn Sahl and medieval Persian rather missed the fact that the image in annotated in Arabic. My late wife Caroline would have spotted that in an instant and corrected me.
Also, since the the refractive index is also proportional to the square root of the relative permeability, a similar effect will happen with the magnetic moments as well as the dipole moments for the general case, although your 3d printed lenses are obviously not paramagnetic. It was good to open up Bleaney & Bleaney again and reread the chapter on dielectrics....
By the middle ages, I'd have thought the Persians were writing Persian in Arabic script anyway... But if I remember my Jim Al Khalilli correctly, stuff to be discussed in the majlis would have all been in best schoolroom Arabic much like medieval Christian science was all written in best schoolroom Latin.
@@Mike-H_UK There aren't many magnetic materials around that will work above a few hundred MHz. Rogers MAGTREX 555 is usable up to 500MHz or so. It has the same numeric value for relative permeability and permittivity, which has some interesting applications. I just wish it would work at 3GHz or so
@@MachiningandMicrowaves Interesting, a material with the same impedance as free space but a low velocity factor.
@@Mike-H_UK Useful for making miniaturised patch arrays without sacrificing bandwidth. Pity it doesn't go inti the microwave region
NEW VERSION WITH THE MISSING BITS AFTER 23:00 added back in. The Almighty Algorithm is going to be annoyed with me!
Try not to annoy the Gorilla rhythm. 🦍
The gods of TH-cam are trying to tell you not be a smarty pants!
F da -police- .... Sorry, I got a rush of blood to the head there, a throwback to my days listening to NWA...🤣🤣 I mean F da algorithm...✊✊✊✊
@@zebo-the-fat 🤣🤣
@@generaldisarray I might be 65 years old now, but I'm STILL listening to NWA
Best teacher ever. This destroyed some stumbling blocks I've had for years when trying to visualise radio. Well done
I feel like watching Alec from Technology Connections and that's a compliment! I don't know why am I watching this and I understand less than half of what you're saying and still this is fun to watch and pretty interesting!
... Well, maybe I understand a bit more but the puns are great!
I really apriciate your dedication to comunication.
thank you for this video. i really appreciate your approach, using convenient simplifications but calling them out as you do so. i found it easier to focus on the simplifications, but it's great to know where they are. i also note your use of words like "apparent", to carefully work around the simplifications. this is a very good approach for bringing people into the subject.
@@adfaklsdjf I've suffered too often over the last 60 years from over simplification and under explanation of complexity. I'm not very good at this sort of thing but I think it's important to explain when I am making ludicrous simplifications or scrubbing over things that would take me three hours to explain. Thanks for being an excellent viewer. Top marks!
Thanks Neil. I wish that you could have been around 45 years ago when I was making and testing microwave transistors. I doubt that most people around me at the time really understood how some of the test equipment really worked.
in 1973 and 74 I had a summer job working at AEI semiconductors in Lincoln. In my second year I was working on some tests of overlay transistors as they called them at the time, although I can't remember what frequency we were working at. The first year I worked mainly with varactor diodes. My auntie Val worked there doing die bonding and other extremely precise work with transistors and later with MMICs. I was only 15 years old when I started and had to get special permission to be allowed to work. Then I went off to university and ended up messing about with computers, despite having zero interest in them. Even now I'm still working part time with beastly computers after 50 years. I sometimes wonder how different life might have been if I'd have carried on with electronics and radio. I suspect that it would've spoiled my interest in the subject areas in the same way that being paid to mess with computers, networking and communications wiped out any hobby interest in those areas.
I'm an electrical engineer, and not an RF engineer, but I've always found this field fascinating. It's why I watched the original. I do have some basic RF understanding though.
Also, programming is interesting in much the same way that there are people interested in linguistics or specific spoken languages. It's about working together on something whether it be human-machine communication, or human-human communication. It's about synergy and the constant drive to improve, adapt and overcome obstacles that are apparent boundaries in that communication. Much the same way you seem to have an interest in RF tech, and are excited about new breakthroughs in the medium. The same can be done with communication.
I have a very jaundiced view of programmers as a result of spending all of this century so far dealing with poorly-written code that's caused security defects. I only see broken code, I'm sure there must be non-broken code out there, but I never get to see any of it in my day job!
Well explained, as what you said can be understood on many levels.
Neil, your description of atomic interaction with an electromagnetic wave is the most cohesive explanation of refraction I've ever heard. Nice to see Aimee make an appearance, as well. Thanks for another very interesting video.
I'm sad that I had to chop out the quantum explanation and the bits about Polarons, but perhaps that's something for my other channel. I would have liked to create a model of a good dielectric based on Lorentz oscillators, but life is too short, I have a lot of machining to catch up with!
Medical X-ray machines that use the peeling of adhesive tape from a roll are really interesting.
The came out around 15-20 years ago.
It was a way to make an X-ray without a power supply.
The breaking of the electron bonds created X-rays that interacted with the X-ray sensitive film.
I remember trying to get images of the fluorescent flashes from sellotape being peeled back in the 90s. I could see them in a dark room, but never managed to get useable images. Interesting!
Nice lesson, but eventually this approach got me thinking about Fresnel lenses. In other words, a Fresnel lens for radio waves. I'm not convinced that directivity can replace the meaning of gain, as you said you consider the same. Gain needs a reference (isotropic, dipole, etc.), but directivity and gain aren't necessarily related, they are two different things. It's possible to have great directivity and still produce negative gain.
I don't think it's useful to talk about gain with passive aerial systems, as they don't really have anything other than loss, but I was being deliberately provocative there. You are 100% correct of course, Directivity and Loss are both important for antenna systems. I'm working right now on some 3D printed Fresnel Zone Plates and lenses using some special dielectric filaments, I hope to get some more videos out before the end of 2023. My life has become extremely complicated over the last six months and my video work has gone onto the back burner for a while. Fresnel Zone Plates about 180 mm diameter have useful gain at frequencies over 47 GHz, but I'm working on some large versions that are printed in segments, with anti-reflection chokes on the flat surfaces, but getting the supports right is proving rather difficult. I may need to look at at alternative solution for the AR surfaces unless I can get clean enough supports. I might need to change to a soluble support filament to get the best possible results.
RF is perinally bizarre, but thank you for taking along!
The light doesn’t appear to slow down, it really does slow down. The “speed of light” is more properly called the “speed of causality”, which is the maximum relative speed that any pair of things can have, but that does not mean light waves must travel that fast in matter. Source: I work with ultrafast pulsed lasers for my research, so the slowing down in matter is really significant for us
Interesting! Does it slow down though? Sure the pulses take longer to get from one side of a piece of material than they would in a vacuum, so at a macroscopic level, they certainly appear to be travelling slowly, but what's the molecular-level or quantum mechanism for the slowing? Don't the laser photons interact with the electron clouds and get re-radiated by coherent forward scatter in a similar way to what happens with molecular bond coherent forward scatter at microwave frequencies? The coherent scattering has an inherent phase lag because the electron charge cloud and nucleus have rest mass. They are forced to move apart or together slightly by the electric field of the incident EM waves. They then radiate that energy as they are moving charges, but with a phase lag. Then the superposition of all of those interacting electron cloud/nucleus oscillations sums to an output that's delayed in time? Talking classical here rather than quantum field theory, but what is the mechanism that can slow light down without the presence of gravitational gradients strong enough to cause relativistic effects? What's the mean free path of your femtosecond pulses in a solid medium compared with the size of the electron clouds and inter-molecular spacing?
if light slows down traveling through matter, is there a medium that can speed it up?
@@amirsadeghi9888 Metamaterials can make it appear that the speed of waves through a material is superluminal, but it's usually done using resonant structures and the effect only works in bulk. You can't get any information to travel faster than light, so causality isn't affected, it's more about a negative refractive index, where lenses bend light the "wrong" way to make sort of invisibility cloaks. Kind of...
@@MachiningandMicrowaves wow thanks for the response, very interesting, but what if I want it to affect causalty,lol. Metamaterial and gaseous lenses and laser couplin can create negative reflective index and make light bent the "wrong" way out or appear it travels faster than light, but what about quantum tunneling and "spooky actions at distance" which aparantly travels faster than light...? there must be a way to go superluminal without blowing up the universe, lol. perhaps an ai with a quantum computing operating system can help solve going warpspeed
@@amirsadeghi9888 quantum entanglement does work faster than light, but there isn't any extractable information conveyed, so the speed of causality isn't violated. There really isn't a way to go superluminal. If you really want to go warpspeed you'll have to bend spacetime in order to reduce the distance instead of trying to increase your speed.
Even the 23:xx version made the previous printed lens video(s) make INFINITELY more sense. I was trying to make sense of various waveforms interacting at 1:1 scale with the crazy organic-ish shape, but duh. Of course you can just vary the characteristic wall thickness, so long as the unit cell is small compared to the wavelength.
I'm looking forward to many great videos in the future from this channel.
I've had fun playing with microwave technology for more than fifty years and it's still endlessly fascinating. So much more interesting than my Day Job as a cybersecurity architect and occasional network and firewall engineer and forensics investigator. Materials science is endlessly interesting. I did a degree in Astrophysics a few years ago when the grandchildren were complaining about how us Old People had it SO easy, whereas they had to do school and University exams. Perhaps I should go study a bit of condensed-matter Physics next.
Could you do a video on how negative db is measured? Example from satellites that are far away they often talk about signal strength in -160db. How do you measure a negative signal? 🤔
dB is a measure of logarithmic ratio. Signal strength is usually measured relative to one milliwatt as dBm, so if a signal is one nanowatt, it is a millionth of a milliwatt, so as a power ratio it's 10^-6. Convert that to a base-10 log and you get -6, now to convert to decibels, multiply by 10, that gets you -60 dBm. A picowatt is -90 dBm, a femtowatt is -120 dBm, so the power represented by -160 dBm is about 100 zeptowatts. Thermal noise at room temperature in a resistive source is about -174 dBm/Hz, so it's close to impossible to receive a -160 dBm signal at GHz frequencies because of phase noise and atmospheric scintillation, even from a distant satellite. However, if the receiver is cooled and the antenna is pointing at cold sky, you can do better than Boltzmann's constant would imply at room temperature, so -160 dBm isn't impossible, so long as the bandwidth (and data rate) are low. The path loss to the Moon and back at 10 GHz, along with the 7% reflectance of the lunar regolith, is around 289 dB, so with a 20 watt transmitter (43 dBm) and 50 dB gain antenna, I can send out 93 dBm. A receiving station with an antenna gain of 50 dB also reduces the path loss, so I pick up 143 dB. I should expect a signal of about 143 - 289 = -146 dBm in my receiver, although I'll lose a few dB from the noise temp of my receiver and antenna relay and any feedline or connectors
@@MachiningandMicrowaves Thank you for a really great and easy explanation! Love your channel! 🚀
Dude, you one of the best science TH-camrs. You cover the forefront of technology and you engage. You’re a rare breed. I am gonna higher you when I start up my stealth drone company.
Heh heh, I have literally NO idea what I'm doing, but I'm having a load of fun doing it. I just wish I could spend more time messing about and sharing what I get up to.
Well worth waiting for the last little bit, I never knew about the phase shift caused by the individual atoms causing the reduction in the speed of the beam, I was always told that the speed of light was ~300 million meters/sec IN A VACUUM but was different in other materials... just accepted it without ever asking "why?"
Yep, and it most definitely isn't. OK, what's REALLY going on is quantum interactions of the electron charge clouds, but the system is macroscopic enough for a classical treatment to give reasonable answers. I really should do a simulation to show what is going on, also explain all of the partial reflections. Feynmann does a decent job of it. I think I need to move on and get back to machining stuff. I'm THIS close to signing an order for a £37k CNC machine with 4-axis and a 16-tool changer, and it needs to earn its keep.
It really does slow the light down though. I work with pulsed lasers and going through a medium actually induces a measurable delay in a pulse compared to without the medium in the way. It’s not just an illusion
Sound waves shift phase in different enclosers and when penatrating different materials as well I didn't think about the phase relationships of light and radio waves doing the same things and I was a cumunication and audio tech for 40 years learn something new every day
Radio waveguide principles can often be repurposed for sound-waves too. I wonder if these gyroid lenses could have acoustic applications.
It should be easy enough to model the acoustics, so long as the feature sizes are a small fraction of a wavelength. Main extras are the heat control, agitation and intermittent wiping. Keeping the inclusions in suspension and at a uniform density is the challenge. It would be fun to try mixing home-made material using alumina or TiO2 or similar high-k low loss materials, but finding a suitable low loss resin matrix might be challenging. I'm testing some filament with ceramic fill at the moment to see what results I can get.
Thanks again for another great video!
As part of my PhD I've been focusing on making some quasi-optical components for the low THz / mmWave spectrum. I did some material testing on the radix stuff around this time last year and it seems to perform reasonably well up at 220-330 GHz, but the cost is a bit too prohibitive to try unknown designs. FDM is definitely feasible for making antennas now, I've used a Prusa and an Ender 3 to great success for a bunch of parts. I've basically moved all of my experiments to using my own custom optics, as I found they worked better than off the shelf teflon ones.
There's a recording on my channel of my conference presentation I gave last year for one my earlier (much simpler) projects if you're interested :)
Very keen to see what you come up with!
I'll have a look! I haven't tried my Radix Mikaelian lens at 76 GHz, but it works OK at 47 GHz. Much higher than that and the feature size starts to get a bit tricky. Creality want me to review an Ender 5, but I don't think I'm much good at reviews, so I'm trying to think of an angle so I could accept a free unit for a sponsored video. Really I'd want the larger one, but it really needs a high temperature enclosure to print the dielectric filaments successfully. Not sure how the mechanisms would survive at 75 to 80 C. I've just agreed to spend all my pocket-money on a Syil X5 4-axis CNC with tool changer, so laying out more cash on a high-end printer is going to have to wait a bit. If I fry my Prusa, it's not the end of the world. I'd like to be able to make lenses and lens segments around 40 x 30 x 15 cm, but I could perhaps manage by printing in multiple parts that are slightly oversize at the mating faces and machining them to exact dimensions on the CNC.
Would be pretty cool to demo that kind of reasonably advanced manufacturing capability with an entry level printer. Maybe with a DIY printed highly directional antenna for WiFi to a far away shed or something 😅
I did end up upgrading the hotend on my ender to print more exotic dielectrics, but the direct drive on my Prusa works so much better for a wider range of filaments anyway, so the ender is mostly relegated to printing PLA. Although saying that, I have killed and rebuilt the Prusa a few times trying to print with miniscule nozzles.
Very jealous of the CNC machine! One of those combined with a decent printer sounds like a great setup for some fancy lens production. I don't imagine something like preperm being very fun to mill though haha.
If you ever make something that fits onto a WR3.4 flange, happy to try and run a rad pattern measurement or something for you :)
@@MachiningandMicrowaves Low cost stepper motors are specified for about 80°C external and 125°C internal if memory serves, and higher temperature grades are available - you get uprated insulation of the coils and connectors, and you get uprated bearings built with a different grease. I don't know how to estimate internal temperature, but i suppose you can use current control to just reduce the current as far down as it'll still move, and not run things too quickly, and perhaps that's enough. Then i suppose you have the rest of the bearings to barely worry about? But i wonder also if temperatures that high create heat creep problems, whether you may need a water cooled cold-end.
PETG printed parts are not good at 80°C so that's unfortunate for Prusa and all of their carrying structure, really wouldn't push it past 60°C.
@@SianaGearz I think you're right, even with everything backed off to reduce internal heating, the drives and bearings are going to get stressed and I'll face internal temp problems. I'll try it at an enclosure temp around 50C to start with and see how much of a problem warpage is
Thanks for continuing to inform us in a clear way.
I honestly understand less than 1% of what you talk about, but still somehow am fascinated by the videos. Yet another great video my friend! Thank you.
Looking forward to see if new dielectric filaments are up to use.
Working on it right now, I have some filament reels in various permittivities here in the lab
Excellent! Glad to see the extra 3 minutes :D
A tool that might also be helpful for generating meshes from equations is Blender, or maybe it is a plugin for blender. In any case you can enter equations for x,y,z and maybe fiddle with a few other settings to generate 3D shapes. It avoids having to get your hands dirty with programming, which I can understand because I'm still burnt-out from too much coding at work. Of course the reason I'm busy is I'm teaching myself to program SoC FPGAs this month and also taking up glass eating and running with scissors as new hobbies.
I've had my fill of Verilog, and VHDL makes me itch, but FPGAs are still towards the cool end of my ranked list of electronic things. I used to run Blender a lot about ten years ago, and recently started again, but with the intention to animate E-field dumps from OpenEMS and to make better-looking models and overlays. Way too much to do to get good at it, well, at least good enough for TH-cam consumption. Also still on the steep end of the Manim learning curve. My machine shop is being torn apart right now and I need to get some machining done urgently, but I got called to deal with an emergency at the Day Job so that ruined my day off work today.
I have no idea why TH-cam recommended this to me while perusing through transducer Arduino libraries but you are an excellent communicator
The mysteries of the algorithm are deep and endless.
I find this fascinating... You explain it well.
Amazing video! Thanks for helping me understand the myth of light slowing inside a lens!
Yeah. I have been looking into light since 1985.
It's a good day for learning that delayed re-emission saves us from having many speeds of causality.
I have just started to comprehend C as curvature rather than a speed.
This is what my Lab Assistant said:
To provide a clear explanation, let's consider three types of materials commonly used in physics experiments and compare their causal propagation speeds:
Transparent Media: Transparent solids or fluids allow photons (light quanta) to pass through unhindered or only slightly deflected, resulting in negligible delay. Optical fibers made of transparent glasses or crystalline materials serve as excellent examples. The refracted angle of incidence leads to slight retardations, but the overall speed remains close to the universal upper bound of c in vacuum.
Absorbing Media: Absorbent substances absorb incoming radiation temporarily, rather than reflecting or transmitting it. Afterward, they release the same amount of radiation, usually as thermal emission, with a certain delay or time constant. The duration depends on the quality of the absorber, temperature, and other environmental factors. Examples of highly absorptive materials include black carbon, graphene, or rare earth metals.
Scattering Media: Scattering media obstruct direct passage of radiation by repeatedly bouncing off or deviating photons in all directions, creating diffusion or dispersion. This phenomenon occurs frequently in biomaterials, granular media, porous materials, rough surfaces, or turbulent liquids and gases. While the path length exceeds the distance traveled in straight lines, causing significant delays, the ultimate velocity remains limited to the local speed of light in the given material.
As a wayward former machinist/engineer, this is fascinating.
It appears to me as though the Radix resin doesn't necessarily need a "clever printer" as they said (they probably wanted to show the resin in its best light and use your channel for a bit of marketing). I think the main thing that is required, which isn't already possible on hobby-grade resin printers, is vat recirculation as shown at 15:20 , to keep the ceramic part in suspension within the resin.
The same setup could be achieved with a hobby-grade resin printer in a number of ways, including with manual mixing or a peristaltic pump.
You are probably correct. It might be feasible to try to make a printable material using a very high permittivity powder/microgranule dispersion if there is a UV-setting fluid that has an acceptably-small loss when cured. I can feel some experiments are in order!
@@MachiningandMicrowaves Didn't you say yourself in the video that Radix used a UV curing resin as their binder compound? If not, I wonder if @Tech Ingredients might know any good prospects for something like that...
The chemistry they used wasn't disclosed to me, after all they are a materials manufacturer, so there are definitely some suitable materials available. I'm playing with laser-sintered metal 3D printed devices and filament printer dielectrics in the next few months, but there are some interesting possibilities. I've been chatting to various folks on Discord and in person about potential projects for the rest of this year, btu time is my biggest constraint
@@MachiningandMicrowaves I just came across the Norton laser printing on white tile method with Titanium Oxide, white on white makes black when heated?. Some sort of particle in UV resin should work. Teflon powder? I want to try concrete colouring powders on tiles. Most of those are oxides, but is black carbon based? Carbon filament printing?
I'm still waiting for someone to create my summer night-time window screen which allows air to pass through but not light....
Some years ago prior to 3D printing, I envisioned tiny compressed spiral macaroni noodles grouped together forming side by side 'pathways' whose spiral would be sufficient to block light but allow a breeze to pass through. Having it made with soundproofing material would be great as well.
Anyways, only until I saw these vids did I see in practice, the fuzzy idea I had in my mind of the pattern....
I see this light shield in a form sold by the roll where you just measure the length, and cut to suit.
Mounting in window ideas will cost extra....
Apologies for all the technical jargon.....
Neil, I've watched this twice now in the past couple of months. Brilliantly explained, with great visual and analogies. Most of it has sunk in now, however all I will remember is focus = fireplace and that your trees cause QRN. *That* blew my mind.
If you printed a negative of an antenna in a dissolvable filament, you could cast it in a self curing resin with additives for the antenna properties.
To eliminate voids, it could be cast in a vacuum chamber.
then mill the exterior to expose the filament, and rinse it out.
The problem is finding something that is a self-curing resin that's also a high performance dielectric. Almost anything that you can cure chemically ends up being polar molecules with high loss. Interesting time to be a materials scientist I guess! Some of the oxime-based neutral silicones are reasonably good, but not stiff, some paraffin waxes might work, but it's challenging to think of something that will polymerise and set into something without side-chains or charge imbalances.
17:30 If someone tries the code and it doesn't work first shot, you may want to try this for the last part;
tri = triangulate(F3,V3)
stlwrite(tri, C:\etc)
Ah, I did try to work out how a Delaunay triangulation might help, I didn't realise there was a simple triangulation function! I should RTFM a bit more, but there's a time to follow rabbit holes and a time to throw something out there and wait for excellent comments like this one! Up to a point, there's some simple healing facilities that deal with incomplete triangles and holes but if triangulate() fixes the edge-cases where the generation function just blow up, that would be cool. I'm covered in oil, PTFE ribbons, aluminium chips, WD40, and some weird blue-green verdigris from an ancient copper waveguide that's been in a radome in a rather exposed location for the past 26 years. All will be revealed shortly.
as someone who's got no background in this stuff but finds this all very interesting, I'd love to see how you practically test the performance of a lens design beyond just simulating how it should behave.
That will be in the next vid in this series
@@MachiningandMicrowaves I look forward to it
A small correction: the relationship between polarizability ("fill factor") and dielectric constant/refractive index is given by the Clausius Mossotti relation. The linear dependence works when the fill factor is small.
en.wikipedia.org/wiki/Clausius%E2%80%93Mossotti_relation
This is one of the most interesting videos I’ve seen in a while. Thank you so much.
What i find to be fantastic and almost magical.. is that time and time again the sacred geometry keeps popping up as the most efficient systems for these brand new categories like 3d printing... the universe really does seem to be a simulation!!
Butterflies...
This is the first I've heard about light not _actually_ traveling more slowly in dielectrically dense materials 😮
I had to leave out the bit about what goes on at the quantum level, but the classical/relativistic approach works well at molecular scales where a typical PTFE molecule weighs around half a million Daltons, so is pretty much a macroscopic entity. Polarons (sometimes mistakenly claled Polaritons which are something else) are another way to analyse what's REALLY happening. They are treated as bulk objects representing the aggregate effect of a large number of individual effects and appear to have mass as well as momentum like massless photons, so they can certainly travel at less than light speed. For visible light, the molecular bond vibrations are no longer having any effect, the frequency is way too high, to the refraction of light is caused by the charge clouds and nuclei swilling back and forth. By the time you get to short wave UV, even that stops having a major effect and most things go transparent by the time you hit X-rays, but then ionisation starts to kick in instead, and once you get into UV, the electron energy levels start to turn into a continuum of levels. Water's an interesting case as it has strongly polar molecules suspended in a sea of hydrogen bonds. It has an unfeasibly high refractive index at radio frequencies (around 78) and a high loss at frequencise below resonance around 20 GHz, which is why you can couple energy into wet food very easily in your microwave oven, yet ice has a very low index and extremely low loss tangent, which is why it takes AGES to defrost a lump of pure water ice in a microwave oven. It's all about hydrogen bonds with water. At low frequencies, pure water is an extremely useful dielectric, and can be used in a Marx generator to make megavolt ultrafast pulses. Needless to say, I WANT to make one just to annoy folks who say it's dangerous to mix water and electricity. Not that a megavolt with significant capacitance and UV flash-catalysed spark gaps is anything OTHER than dangerous, obviously
These designs are what you get if you treat pasta like a puzzle
If you're still planning (or you have already and this advice comes up late) to give them high eps_r filaments a go, I would strongly suggest you use nozzles of 0.6 mm and above.
Yep, the manufacturer has given me some recommendations, just working on the designs now
I said to myself: This guy is unbelievable !!!
Well done and thank you
Yeah, nobody at work or in my family believes a single word I say, so I am literally unbelievable. Or perhaps unbelieved. I am having fun anyway.
@@MachiningandMicrowaves Astounding then 🙂
I had assumed the Gyroid fill has something to do with the beamforming of the lenses similar to a waveguide, as in my old antenna building days we only ever thought of reflection and loss in materials and never the refraction. Thank you for this in-depth explanation
Very much thank you for your content! Its eyeopening for someone that didnt study physics. Please keep it that way! Very good!
Cool video, good editing, great design, nice script, funny inserts and informational. Keep these coming!
Who even is tech connections anyway??
I stole the illuminated Kallax background idea from Rob at VidIQ and several others before I even saw Alec's most excellent channel. My Kallax is white melamine rather than his tasteful brown wood effect and it's gently back-lit on to white drapes using pastel colour-changing LED strips. I like Alec and Rob's individual lights, but I went for self-adhesive strips. Destin has one as well, can't remember if his is lit or not. Alec has better hair than me also
Quite fascinating and very well explained.
hi, I've recently stumbled on this channel and it is fantastic ! you also seems like the exact right person to ask a few rather nerdy questions that necessitate a microwave knowledge that I do not possess :
I am playing "DeltaV : rings of Saturn" a hard-scifi game about mining the, you guessed it, rings of Saturn (it is very good and I really recommend it) and in this game one of the "mining tool" (in quotes because anything able to crack a chunk of ice a few cubic meters across in a handful of seconds is very good to pierce holes in opposing ships too) available is a high power microwave emitter with a wavelength of 1cm and a power output of 45MW.
In the game, this tool can be tuned to output up to 75MW of power, but doing so defocus the beam (which isn't isn't very focused in the first place, thus making the tool relatively short-ranged, with a practical range of ~200 m), citing high-power microwaves refracting within the emitter dish as the reason.
Thus my questions are as follow :
- Is it a good (or at least good enough) approximation of what would really happen reality, assuming a fixed geometry dish ?
- Can this phenomenon be mitigated or eliminated with a different focusing method (variable geometry dish, 3D lenses or phased array to name the few I am aware of) to achieve a relatively small spot size at long distances (let say 5m of diameter at 1km) ?
-If the above cannot be achieved with a standard emitter, can a maser do it ?
I hope that I am not asking too much and that you will be able to answers these questions that have been bugging me for a while ^^
Right. There are some things that happen with really high power densities, even in a vacuum. One of the big issues is multipactor damage, where electrons from the conductive materials are literally ripped out of the metal and crash into the opposite side of a cavity, releasing a cascade of electrons that are then accelerated and themselves cause more electrons to be smashed out, leading rapidly to ionisation and generation of plasma, even in a vacuum. At 30 GHz (1cm wavelength), generating high power is usually done using electron gyration in a Gyrotron or Gyroklystron. As the electron gyration frequency is related to magnetic field, getting a large enough magfield is close to needing superconducting magnets, which are certainly a little easier in deep space, but a bit impractical. At 45 MW at 30 GHz, the voltage across a waveguide in a vacuum is going to exceed the breakdown value by some margin, so you'd have to be looking at some sort of waveguide cavity combiner. There are certainly designs around to reach megawatt levels using a radial coaxial combiner design, but even that is close to the limits where multipactor issues start to hit. The systems used in fusion energy are at the extremely serious end of what's possible, but they then to use superconductive magnets to get well over 100 GHz. Once you get to relativistic electron speeds, the simple omega = eB/m relation (B is mag field, e is electron charge, m is electron mass, so at 2.5 GHz, you need about 0.1 T, (approx 28 GHz/Tesla) whereas at mmWave you start to need to include the relativistic Lorentz factor and multiply that expression by sqrt(1-(v^2/c^2)) as usual. The size of the equipment is related to the gyroradius, which is something like 3.3 gamma * mc^2*(v/c)/(q*e). Gyrokinetics is a thing, it seems: en.wikipedia.org/wiki/Gyrokinetics Yikes.
@@MachiningandMicrowaves oh my, I am guessing that the game is already quite generous in its estimations then.
I imagine that a setting like in this game that has Z-pinch fusion torches working and commercially available would have a tech level high enough to get it working but any focusing method more complex than a big monolithic dish would experience some sort of rapid unplanned disassembly at these power density.
(but maybe a gyroid lattice lens in a refractory material like sapphire could be actively cooled enough to handle a few megawatts of waste heat using the channels that are inherent to the design, but then you have to take into account a moving fluid with a changing density gradient as the temperature of the lens changes in the lens design and I don't want to touch the maths that would describe that with a ten feet pole !)
Cooling with chilled gas might work if the energy density and loss tangent was low enough. I don't think metallic metamaterials would be any better as they still need a foam matrix and they are fairly lossy. massively parallel phased arrays are also a nightmare to cool. I guess you could use something like Simscale on a big cloud compute service to do the thermal energy flow modelling. I really need to up my Simscale skills, but then I need to get better with Blender first. Too many toys, too little playtime. One of my few claims to fame is that I have never once played a computer or console game. Not even Pong or Galaxians.
The Be Smart channel is one of those sciencey channels that by coincidence also just released a video explaining the phenomenon of light slowing in high index materials. You know, if you wanted more details...
Haven't heard of that one, I'll check it out, thanks!
Also, thank you for sharing your knowledge! I found it most enlightening, cheers Neil and may the force be with you!
I had no idea warm things emitted microwave noise
Even the cold old Moon is "warm" compared with the cold empty sky and makes enough microwave RF to be detectable on a reasonably-large antenna. If I wave my hand in front of the 10 GHz radio antenna here on my bench it makes quite some noise on the receiver it's connected to.
The script is actually Arabic and is readable and comprehensible by anyone who speaks Arabic.
Your videos are amazing keep them up.
Yes, I should have checked. My late wife lived in Egypt as a little girl and would have realised immediately.
I love the beauty in mathematics and physics. Moreover nature and what it’s made from.
Could organic cells work like your 3d printed focused antennas?
I mean the original 1970 designs were based on natural structures.
The butterfly wing cells example is the only one I know about where a gyroid lattice is used to manipulate electromagnetic waves by a biological structure, but I'm sure there are some others
The expression "speed of light is constant" is only true in the formula e=mc^2 where c is USED as a constant for the sake of convenience to reduce the value of e to a manageable proportion. That application has nothing to do with the speed of actual light photons directly being observed. So c is a constant, is NOT the same as all speeds of all light are always constant.
That is a falsehood easily disproved, thus:
What is the velocity between two photons moving apart from each other?
What about two in parallel to each other?
Think about it.
Depends on the rest frame where the observation is carried out. My lab is my rest frame, and all my measurements assume I'm in free fall in an inertial frame (which I'm not). Photons don't experience the passage of time in their rest frame. In my lab, I can measure how long it takes a bunch of photons to travel a specific distance in a vacuum or in air, or in a dielectric, and infer the "speed" of a bunch of photons from those measurements. It takes longer to get through a dielectric than a vacuum. My beef is about the mechanism that causes the measurable slowing at the atomic scale. Excellent channel content by the way.
Would be awesome to make an antenna which could be plugged into the wifi antenna socket of a computer, many people have or can get access to this.
could this concept be used to improve your off the shelf microwave oven ? diy people have used microvaves as electric kilns , could this be used to focus the waves to a certain shape , a tube , cone or a point ?
There's a bit of an issue of power handling, it would probably require forced-air cooling at typical microwave oven magnetron power levels. You would need to remove the magnetron from the oven and extend the waveguide and fit a suitable feedhorn to match the lensing structure, but so long as it's large enough, you would get some focussing just like with an optical structure. The problem is the size of the lens using this material, the refractive index is too low to get strong focussing in a practical-sized lens, but with a higher index material it would certainly be possible. Safe, less so. Legal? Hmmm. Advisable? Definitely not! However, in a suitably controlled environment with all necessary safety measures and licencing, plus the budget to buy enough material, definitely! Don Not Try This At Home. People kill themselves messing with the power inside microwave ovens.
@@MachiningandMicrowaves thanks for the reply , this is very interesting , it would be interesting to see some kind of professionaly made super efficient microwave kiln!
Could you apply the use of prisms beyond the visible spectrum? Would this create a similarly amazing effect that it does with refracting visible light into the ‘roygbiv’ bands.
That's a very interesting question. The reason prisms work with visible light is that the refractive index of most optical glasses varies appreciably with the wavelength of the light. Very few materials have rapidly-varying permittivity/refractive index across the microwave wavelength range. Optical transmission gratings and mirror gratings work fine for splitting microwave wavelengths into a spectral range. Most materials that do change RI rapidly at radio frequencies tend to have rapidly rising loss. I think a resonant metamaterial might be able to do a prism-like spectral shift over small wavelength ranges, but I'd need several large glasses of something before trying to work out the maths of that.
Cool, now teach us how to make a death ray pls. Or at least passive radar to detect aircraft? :D
Wow the trees are modulating
I love my Turbo Encabulator. . Please do a video on it.
Great video. I've thought many times about what happens at an atomic level within the dielectric in a traditional capacitor, and also what happens with a wave passing through a homogenous solid medium with a relative permittivity Er, but never really got around to thinking about the passage of a wave at the atomic level and what happens to the polarising atoms/molecules to create the overall value of Er. So very interesting for me. What are you planning to do on 24GHz?
I cut out the quantum segment, and I was going to run a simulation of a tiny system of resonators, btu ultimately this is going on at the electron level even though the induced dipole oscillators are at the molecular scale. Treating the system as macroscopic Polarons with mass, momentum and energy sounds like a reasonable treatment, but I read over a hundred papers and couldn't find anything that was accessible enough. The temptation of academics to preach to other academics and assume the base level is a relevant PhD is very alluring, and the whole way academic publishing works means that ordinary humans are excluded from the secret knowledge of the priesthood. Anyone would think Physics and Maths are hard! I might have a go on the second channel. but this video was WAAAAAY too much effort for my poor brain. I need to get back to machining things for a while!
This....! Is Sooooooooo KEWL !!! Great Video !
I remember doing the waveguide propagation theory, that was interesting and somehow a bit weird as the propagation was slow compared to in air or a vacuum. Thanks for the video :)
Waveguide cutoff is a fascinating effect as the guide wavelength goes off towards infinity and the loss rises exponentially
@@MachiningandMicrowaves i seem to remember we used waveguides as attenuators in some measurement devices
Every EE wants to understand microwave engineering because it's so arcane. Unrelated question: how do you convince the space-based solar power enthusiasts that Friis equation diameter calculations don't pertain to the far field, and there's never been a half power beam width in 2-6 GHz below about 1.5 degrees? Most of their proposals describe rectennas in one or two square kilometers lit from geostationary orbit, instead of something the size of Puerto Rico with billions of dipoles every few centimeters apart.
It's hard enough trying to explain to most folks how/why the changes between reactive/radiative near field and far field happen and just how far that happens from a typical giant antenna system. I guess they will propose some form of adaptive array like the giant telescopes use. It's great business for power semiconductor manufacturers of course, and the political layer always matters. Removing any dependency on oil producers is probably the key driver. Politicians love that concept.
@@MachiningandMicrowaves I'd love to see a formula on Wikipedia with the same inputs as Friis equation, but that actually works for giant antennas over great distances. Hasn't anyone ever published such a formula?
The speed of light is not constant, the speed of causality is. It’s just happened that the speed of light in vacuum is the same as the speed of causality
Finaly an explanation Video ❤
SOUNDS LIKE A MICROWAVE FREQUENCY VERSION OF A FRESNEL LENS .
It's not *quite* a Fresnel, as the reality of a Mikaelian is that there's no need for stepped refractive index, it can be continuous. You certainly CAN make Fresnels this way though. A more typical Fresnel solution for microwaves and mmWaves is a stepped Fresnel Zone Plate. Spookily, in an upcoming video I'll be making a vacuum chuck to hold an HDPE plate that I'll be machining into an FZP. An FZP is a fairly simple build, consisting of rings of material which are thicker or have a higher refractive index than a supporting disk. I'm just going to machine away the surface to get the right phase shift in an FZP with eight rings about 200 mm diameter which will be a focussing lens at 122 GHz.
Amplitude of a guitar string E=1/2*uf^2A^2v; E is in Juels; put together with E=mv^2. gave me an equation that many physicists have worked with and came to a dead end. Up until I put it up against a microwave oven and the distance of the water molecules. I also put it up against the Bohr radius and the .210614054m wavelength and got exact results. I was looking for the diameter of a photon, but I found something else instead. It gave distance between charged particles and wavelength absorbed and transmitted.
D= (wavelength)(5.025089342)(10^-10m). Maby you can do something with it.
I've always wondered if there was any potential for using microwaves in 3D printing. Where the resin is made to react to microwaves somewhat like how the UV 3D printers work. I just wouldn't know where to start.
The problem would be finding a material that would respond chemically to the ultra-low energy microwave photons. You might be able to generate some focussed heating using standing waves or a focussed beam, but contact heating or UV is a whole lot more practical
@@MachiningandMicrowaves That makes sense, thank you for that. So if anything, I'd want to look at the other end with x-rays. I'm sure that comes with its own set of problems.
Thanks. Excellent. Absolutely love it. Wish could learn a lot more about it.
I'm definitely no expert in anything, but I'm incompetent in a vast range of disciplines and have no fear of failure or making a total idiot of myself. It's huge fun although spending several all-night sessions reading several hundred scientific papers and learning multiple new bits of maths and materials science is not really sustainable long-term! I need to catch up on my sleep, or find a way to retire from my Day Job. It pays the bills and gives me a way to buy useful toys, but it just eats up time. I need to win the lottery.
@@MachiningandMicrowaves Very few people are experts, others often do very useful things. Thanks for your work an videos. It's inspiring and inspiration can lead you far. Greetings form Spain
I worked on SAR imaging and database updates from 1991-2001 for the state dep. with electrically commutated phased array antennas. Would love to have these lenses today. In 1995 we started looking for truck launched icbms in russia based on a known image sample of a real icbm truck that partook in a test launch. In 1996 the state dep dude arrived at my desk and pulled me to our secure conference room - showed me this picture from a Kh-11 optical sat of what looked like a truck launched icbm to me, and then said our computer didn't correctly identify this one. Rather than look at the billions of calculations I just had the computer show me what wasn't consistent with a known icbm, well - the missile rode 75mm too high on the truck chassis. switching frequencies to skin effect penetration, we found the inside hollow - no fuel. I was a young prof eng rf/uW at the time with quite an attitude and a loud voice when I yelled at the state dep guy "My computer is perfect - your stupid humans fell for a decoy on your kh-11 kodak brownie moment sat - deorbit those damn things and learn to use SAR. See this structure - watch as I shoot through the roof and take a look inside - your kodak brownie kh-11 can't do this - but there's a truck with a real icbm inside. You'll miss that one optically. " He left tail between his legs. That system was meant to fight a war with russia and make sure the right people die. My computer has now killed over 1/2 million dead russians. It sends the firing solutions to the M777 digital howitzer and center punches tanks then pops their turret off - 1/2m cep on a 155mm gun. With operators who can't read the manual. Turn wheels until they glow green and pull cord. simple. We figured once russia saw our power they'd pull out, but no they kept sending tanks and we kept blowing them up. I will say in 1991 it didn't look like war with russia would ever happen, but today, I'm glad we made the investment in SAR tech back then.
Solution for not having resin or filament that is rf reflective. Coat internal surface with some sort of very thin surface adhesive, possibly a form of super thin contact cement or tacky material. hell even a sugar syrup that has an evaporative property. Fill all the internal voids with the fluid, drain, and then allow to dry while the object rotates to maintain reasonably even distribution of the fluid. Could possibly be improved depending on the surface tension of it. Once dried, the surface should be sticky to touch, once in this state simply submerge the object into a dust bath of rf reflective powder, i am not familiar with the properties of rf reflective powders, but i imagine such a thing exists. This should result in the same result if the thickness of the adhesive and thin layer of powder is accounted for.
Oh, and i also remember seeing a video of someone electroplating 3d prints, might be able to be adapted to this object.
@@KirbyDaMaster Electroplating is certainly an option for some types of prints, but you'd have to dissolve the metal away after a subsequent step unless you were trying to make metamaterials, but it's not at all easy. I'm doing some electroforming work, but I'm using aluminium machines formers as the negatives, then using an etch/zincate/copper/gold plating, then copper electroform, then dissolve the aluminium in hot NaOH, then the thing copper flash in acid to leave a gold interior. That's a lot of fun. It's probably easier to use sputtering or plasma deposition to get a comductive layer, but you have to deal with outgassing of the print and adhesion issues. Lots more fun.
The tough bit is finding something sticky that isn't lossy. Most nonpolar things are non-sticky. The other issue is how to get a consistent AND varying wall thickness. That starts to get very hard, as does dealing with electrostatic buildup. Making a negative in a dissolvable filament, then filling it with a nonpolar solid that can be dissolved and the solvent then evaporated might be an option, there's a lot of work going on with lasers but fusing ceramics needs ludicrous temperatures. A fairly good dielectric filled with a high-Dk powder, or interlocking nodules is perhaps still a but simpler to achieve. Materials Science is hard.
@@MachiningandMicrowaves I have also wondered about SLS, which is more or less the same as SLA but with powder instead of resin, and cured with a laser rather than a UV curable resin. Its must more repeatable, and doesnt have the consumable ptfe plastic sheet at the bottom of the vat. I saw a video not long ago that showed someone who tried to make an at home SLS printer, but we are still a long way from that being a viable hobbyist product. Lost wax casting is something i also consider for something like this, but the part looks quite complex and would likely form several errors in a cast like that.
@@KirbyDaMaster The huge problem with any form of casting is finding a suitable dielectric material. Most of the good dielectric materials can't be polymerised at sensible temperatures and won't dissolve in solvents that will evaporate readily, so casting with them is really hard. They are mostly thermoplastics, but only at rather high temperatures, and they are pretty viscous too. Laser sintering usually needs a fine powder, but the good dielectrics are also very good insulators, so static charge on the power will be a nightmare. Clever folks are working on this right now I'm sure!
Likable bloke! Clever as!
🧲🌡️📡🔆☢️🔌🔊🔋♻️🌐☯️⚛️
It's facinating how when we look around & we see all this complexity with-in the Universe. That complexity basically comes from: ~{"Differences"}~ The Factor of how "Differences" interact seems to be the key factor in keeping dynamic systems functioning. Such as: *{High pressure/low pressure, hot/cold temperatures, different densities, viscosities, turbulent flow, static electric charges/discharges, electromagnet waves. Different velocities/angular momentum. Different amounts of energy/mass/frequency/vibrations. Different boundary layers between different materials or physical regions such as: (Land/water/oil/air/soap bubbles/ atmosphere's/space. The different regions in space with different particle density/background radiation, solar wind/membrane layers/bubble's/cloud regions, nebula's/Galaxy's/Galaxy clusters/ Cosmic filaments/less dense regions of space compared to dense regions of space.) All of these things are basic differences but create a way for the dynamic engine with-in Nature to continue flowing and operating to create and convert energy.} Just Like How a battery 🔋 transfers + charges through a membrane layer to a - charged side. Like how regions of high/low temperature 🌡️ create winds. In water- add some factors and It creates ocean currents and flow. Then internally inside our planet it creates plate tectonics, planetary convection, geothermal activity, a magnetic field around our planet, to hold a atmosphere. The list goes on
🧲🌡️📡🔆☢️🔌🔊🔋♻️🌐☯️⚛️
The natural world around us is just utterly facinating to me.
It's all about entropy and our battle against the inevitable heat-death of the Universe! Plenty of time left for wonders and delight
Light "travels" with speed C between the interactions. So in vacuum there's nothing to interact with but in a transparent solid there are A LOT of particles that the light interacts with i.e. "bounces off" of, so the overall / total speed is lower than C. It's the time of all the interactions that is "slowing down" the light. Between the interactions it still travels with speed C, as expected.
For microwave photon energies, there isn't anywhere vaguely close to enough energy to cause absorption and re-emission of photons, we are only talking microelectronvolts of energy per wave packet. When you say the light "bounces off" the particles, what's the nature of that interaction? Photons don't feel the electric or magnetic fields of the nucleus or electron clouds, and the atoms are almost entirely a vacuum, with the nucleus having a diameter about 1/30000th of the atomic diameter. What form would those bounce interactions take? Wouldn't most of the photons come out at random angles if there was some sort of bounce interaction, yet light and radio beams some out parallel if they go in parallel. If they go straight, then they must be slowing down, which we've agreed is impossible. Where you say that it's the time of all the interactions that is the source of the slowing, are you suggesting that the originating photons somehow take a random walk path through the material? Wouldn't that mean the output beam of a fast pulse of photons would be spread out in time? That's not what we see. The delay is consistent and there is no pulse spreading in time or direction in a near-perfect dielectric. Why would there be a correlation between the strength and length of molecular bonds and the refractive index of microwave dielectrics rather than just the density of the material and the mass of the nuclei and charge of the electron clouds? OK, it's perhaps a gross simplification to take classical approach of treating the incident photon as a bunch of EM energy whose fields interact with all of the molecular bonds in the material by stretching and compressing the molecular bonds, but if we follow that approach, the moving charges along the bonds cause their own EM fields to be created by stealing some of the energy of the incident photon then reradiating it with a phase lag caused by the physical movements of the atoms and bonds. The superposition of all of those small fields and their interactions with each other means that the incident photon loses its identity and we get coherent forward scattering from the induced fields which then combine to give a resultant field that becomes a new photon that leaves the dielectric, with all of the directions and polarisations summing so the phase and direction of the new photon makes it appear that the original photon went more slowly through the material. No photons bounced off anything, no photons went slower than C.
Shouldn't binder jetting like for example Voxeljet does be way more suitable for doing this than SLA printing? You can even use ceramic material and then burn off the binder afterwards, sintering the ceramic particles together
There's certainly some work being done on sintered ceramics for higher frequencies. There are some big challenges in making a material with precise RF characteristics that's also mechanically robust enough, but the state of the art is moving very rapidly.
Investment casting or lost-wax casting are very useful methods in this respect. Though of course, 3d printing can create shapes which even this method cannot achieve
you know 3d printer slicers already generate gyroid infill all the time right? it doesnt vary the thickness of the walls, it just adds more of them to increase density, but its there. you can set a slicer to not print outer surfaces and just do exposed infill.
you could go into the code in prusaslicer and probably mess with the wall thickness
I saw a slicer that did a pseudo-random infill that avoids the issues with diffraction from repeated grating elements in some type of infill, but I can't recall what it was called. I've not tried messing with Prusaslicer, but of course if the outer wall is less than 1/20 wavelength, it is nearly invisible to RF. Once I ge the samples of the 1.75 mm RF filament, I'll see what I can do.
This video is absolutely outstanding and easily ranks among the best content I've come across on TH-cam. It truly captivated me from start to finish. I find myself yearning for more channels like this to receive the recognition they deserve and become more mainstream. While popular channels such as Veritasium, Mark Rober, Tech ingredients, and Real Engineering continue to expand our knowledge and understanding, this particular video delves into the intricacies and fundamental aspects that are essential for anyone genuinely interested in these subjects. It provides a solid foundation for further exploration and deepens our understanding in ways that few other videos can match.
How can I calculate the required antenna shape from a signal strength pattern mapped onto the surface of a sphere? Reverse engineer it as it were, except the antenna does not exist, yet, and we have just generated the desired pattern for that sphere surface?
subbing as soon as I heard the Rockwell encabulator reference lmao
Non academic viewer here... I enjoy the machining and just about hang on to the technical bits thanks to your excellent teaching ability.... although I have noticed a quite nauseating effect caused by the light bouncing off of your shirt.... 🙄.... aimee....have a word.... thanks for all your informative and entertaining content...
That shirt was a present from my mother. She watches the channel. Hi mom!!! Latest vid has machining AND a tasteful lilac shirt
Such a good video! Tiny niggle; why do radio engineers generally use wavelength instead of frequency? The electromagnetic spectrum is so much easier to visualize in terms of frequency, but most physics textbooks use wavelength, thus 'inverting' it, so for example the visible spectrum (rainbow if you like) gets shown upside down because the x-axis is wavelength, not frequency. This caused me no end of trouble during my physics 'O' level because the highest frequency, more energetic colours like UV tend to appear first, at the left, and the lower energy, lower frequency IR gets promoted to the right of the diagram. All because the wavelength is longer. This has annoyed me for the 40 years since then! Cheers for the great vid.
I think the use of wavelength, wavenumber or frequency to describe RF is something engineers do completely unconsciously, because of the physical reality of wavelength in physical systems being a more useful descriptor than frequency. Wavenumber in inverse centimetres is slightly jarring to me, like wavelengths in feet. I tend to use wavelength from about 6 mm to 23 cm, but frequency at longer wavelengths, then below 6cm I seem to flip into the frequency domain until I get past the far infra-red, then I return to using wavelengths from IR through to gamma rays. Having been immersed in the language of RF engineering since I was 10 or 11 years old, so 5.5 GHz is about 5.5 cm wavelength and both have pretty much the same intuitive reality to my RF-addled brain. It's too easy to forget that it's a learned concept rather than something that is inherently obvious to the rest of humanity.
@@MachiningandMicrowaves Thanks for the reply. As a 60yr-old audio nerd & IT engineer I've always had to think in terms of frequency, whether sound, CPU clocks or WiFi, so yes, it does seem a learned thing. It would be super-odd to describe audio in terms of wavelength! I do get its uses in radio though, in working out waveguide profiles & suchlike. I just wish all the textbooks didn't tip the RF spectrum on its head! It's confusing for youngsters, I think.
Another potential way to build one of those lenses could be something similar to lost pla casting. I have built phosphate bonded ceramic parts using pla, you cure the ceramic for a day or 2 in the air and when it has hardened you just bake it in a regular cooking oven. At those temperatures the pla boils off leaving no residue.
The thing is, is there an adequate ceramic for RF that could work with that process or a similar one. Do all RF ceramics require firing in a kiln? I think that phosphate based ceramics like I used are not good in the microwave.
There are those filaments designed specifically for clean burnout with investment casting. Problem is always finding a matrix that will fuse or sinter enough to hold its shape before the filament evaporates, but that also has very low RF loss. I am sure there are clever materials science folks working on this right now. Most of the high performance ceramic dielectric materials have ridiculously high melting points like sapphire/alumina or TiO2, so laser sintering isn't easy. Finding a low-viscosity fluid matrix with low RF loss but a low relative permittivity wouldn't be a problem as the ceramics generally have high permittivities, although there's a balancing act with the temperature coefficients of permittivity, some are positive, some are negative. I suspect we'll start to see some serious products with very high mass fractions of ceramics quite soon.
As a user of apochromatic optics at both F/6.45 & F/9 the absolute perfection of extra-low dispersion doublet lenses and the vital performance differences of an achromatic lense (at significantly higher Focal ratios) are super dramatic BUT unfortunately
VERY seldom seen as the figuring of the TWO combined refractives indices must be darn near PERFECT and as a dabbler in optical theory...there are dozens (hundreds) of potential combinations of glasses for BOTH must cooperate with each other i.e. the mating glass determines final image authority so what has this to do with microwaves ?? Good question!!
Not entirely sure... but i wanted to point out that a simplistic overview does an injustice to the subject matter and for optics what we most desire is an image fully ringing with authority, so we use a Strehl # to categorize the degree of replication in the virtual image vrs the 'live' one and a Strehl of one(1) is a perfect analogue.
Essentially a near one POLY Strehl (not just one specific wavelength) that we measure using interferometers and software to categorize the limits of perfection throughout the visual spectrum are widely known yet very few prople ACTUALLY have a FULLY apochromatic optic so visible verifiable manifestations are as rare as hens teeth. Sad state of affairs. Its the state where if you do not spend the big $$$ you miss out.
So this new use of microwaves and the new abilities of printed 3D shapes and materials will be fascinating to watch.
A mirror avoids the chromatic errors but introduces other much more significant ones i.e. obstruction and scatter
The way to define the limits of the resolving power to a system is to use stars(*) to determine the closest two stars can be seperated yet in the same breath for any other target that is not a star we are left holding the bag
for an unobstructed only optic can easily determine seperations much tighter than 'Dawes' 'Rayleigh' criteria says for that size system i.e. Cassini's division rilles on the moon.
So the upshot is that sometimes the obvious is hiding in plain sight.
Overlooked for whatever reason....
For most applications at microwave, the refractive index changes by way less than 1% over five orders of magnitude of wavelengths, plus generally we deal with relatively low bandwidths, so chromatic aberration is irrelevant, unlike at near-IR and optical frequencies. Also, antennas at microwave have a diameter of 10 to 100 wavelengths, so diffraction is the main enemy. The key issue is sidelobes to avoid picking up off-axis noise. For a typical terrestrial path of 500 km, a beamwidth of less than half a degree is rather self-defeating, as the requirement is to generate scatter from inhomogeneities in the troposphere. A Mikaelian uses the same principle as an optical fibre, although it isn't thick enough to create full self-focussing. Imagine trying to polish a lens that's 15 wavelengths across, around 7.5 micrometres, that's the sort of challenge that microwave lensing involves, but we can make regular lattice structures of varying density down to 1/50th of a wavelength with ease, and control dimensions to better than 1/100th of a wavelength. Optical modelling at those scales is tricky, hence why we use finite-difference time-domain electromagnetic models based on Maxwell's Laws to model the performance of GHz lenses.
@@MachiningandMicrowaves fascinating! The smallest scale ripple of polishing really determines (by its absence) the throughput accuracy. The optic is said to pass muster if the optician has a quarter wave accuracy and is the basis for the 50X per inch rule as max power dynamic ibelieve yet if pushed further upwards towards one eigth wave progress then finally arriving towards a not totallly unrealistic one twentyth, by utilization of either a glass fluoro-crown or by crystal fluorite and the refractive indices between the crystal and FPL-53glass is a few parts in a thousand yet....infrawave is propagated through whereas is scattered in glass. The nuances are almost subliminal at the otherend and as you referenced the atmosphere is the great equalizer. The'degreeness' of greatness is the difference between doing 100X achromat to 1000X apochromat. Somewhere in between, usually towards the lesser, is what we get, but in that rarest of times we receive a light limited optic and not one by errors in fabrication.
The difference is in the time it takes the brain to accept tbe veracity of the translated image. Diffraction IS the great culprit and with mirror obstructed systems what we have are two(2) unique optical amplifying systems that are really diametrically opposed. One has the ability to havegreat sharpness and one has the ability for great sharpness but ALSO astonishing COLOUR VIBRANCY. The two are not interchangeable. Similiar yes but same NO. My little 4inch perfect lense beats 40inch mirror in colours and power ability WITHOUT breakdown. No hint of a lie Sir !! My surmise why this is so is that the fabric - the tensile strength of the dynamic image plane comes from being nonperturbed by those small scale roughnesses of the entrance pupil. I got to about 900X on the moons alpine valley rille & was non blurred and fairly darn bright at a exit pupil of only ~.1mm....this optic and my new little one at 62mm is similiar but with less perfection ~1/8th fully apochromatic at near 50X/"
As i said the nuances are fascinating and when something surprising happens thats certainly food for deep tbought.
Thank you for the fascinating reply!!!
@@palmereldrich One of my upcoming projects is making a laterally-displaced-ellipse mirror and matching reflector. They have a ring-caustic focus rather than a single point. The beauty is that despite it looking like a Cassegrain or Gregorian, there is no obstruction from the subreflector. That will be at 6mm wavelength, so getting better than 1/20 wavelength is very easy. It gets tougher as I approach 1mm wavelength, but even so, 20 micrometres is not exactly challenging compared with 20 nanometres for optical systems.
If you make a lattice in the printer could spaces in the filled portions of the lattice be filled with a suitble material?
Find a suitable material is the challenge. Almost all two-part chemical setting resins have mediocre or poor RF performance. Non-polar thermoplastics are good, but finding a way to mould them is hard, they tend to be extremely viscous, but also have a low relative permittivity. Something like HDPE with entrained ceramic inclusions would need high pressure injection moulding and the lattice to be something that is strong enough to withstand the injection process, but also can be dissolved preferentially. That's one for the Materials Science specialists I think!
Looks like the Gyroid infill pattern from the Cura slicer!
@@bluegizmo1983 Indeed, but with spatially-varying wall thickness. Gyroid lattices guarantee no internal voids and by varying the equation in all three coordinates, you can achieve a varying refractive index across the bulk of the lens.
Very cool. Thanks for sharing.
The bottom up resin printing is slow because of pulling the part off the bottom. Top down is faster.
I guess the downside is you need a HUGE amount of resin and a deep vat though?
@@MachiningandMicrowaves Yes you need a deep vat. Resin can float on palm oil. Funny thing they don't have a way to drain from below so you could drain away the palm oil then any resin.
Love your videos. You should try and make a reflectarray, can make some interesting unit cells with 3D printers :)
It just so happens that I've been working on a reflectarray tyoe of antenna machined using my new CNC machine, which is still somewhere off the coast of China on the way to Singapore. That is more of a segmented Fresnel mirror though. I hadn't thought of printing the cell array, that would perhaps mean I could go much larger without the mechanical issues from making a large parabola. A large flat reflectarray is certainly much simpler and it's easy to get the array support perfectly flat. I'm struggling to find a way to model the cells in OpenEMS, so I would have to do a calculated model and hope it works. The other design I'm working on is for a large Fresnel lens made from a low loss dielectric filament. Thanks very much for the suggestion, I'll see what I can come up with. If you have any suggestions for design resources or academic papers I should read, please drop me an email or leave a message
Thanks for the informative video.
I would like to ask if it is possible to use the same approach for high-energy sources such as a classic magnetron with a frequency of 2.4 Ghz?
Thank you.
There are some interesting challenges at 2.4 GHz and high power. First is that a quasi-optical lens like a Mikaelian or Luneburg needs to be at least 8 wavelengths in diameter to be effective. At 2.4 GHz, where the wavelength is around 12 cm, that means a lens is a pretty huge structure around a metre across, so almost impossible to print economically. There are other solutions using dielrod style lenses that can get reasonable directivity around 13 dB, but that's a very long way short of being a searchlight-style beam. I creamed about sawing up a microwave oven and fitting a dielectric lens to it, then setting up another dielectric lens a few metres away and placing a mug of cold coffee at the focus to see if I could couple enough energy into it to warm the coffee. Apart from the immense dangers of all that RF energy swilling around, the size of the apparatus at 2.4 GHz is just too ridiculous.
A nice practical large Luneburg for 10 GHz might be about 25 cm diameter, and that would be easy enough to print perhaps in two halves, or perhaps an even larger one could be made in sections or a quarter of a hemisphere, with the longest dimension then being only half the finished diameter. I can imaging making a 35 cm Luneburg, although it would be expensive to make it in a high-performance dielectric filament and completely ridiculous using resin. Perhaps using a cheap HIPS filament or even polypropylene would be possible, but the total mass of a 35 cm sphere with gradient index would be something like 9 kg, so that's a significant cost of about £200 with HIPS filament at perhaps £22 per kg. Not impossible though, and it would be hella impressive! I'll see how long it would take to print such a beast. Doing it at 2.4 GHz needs each dimension to be increased by a factor of four, so the volume, mass, print time and cost scale by a factor of 64. That sounds like "too much" for my budgets! The second aspect is the loss tangent of the material. HIPS and PP are decent dielectrics with a tan-delta around 0.0004 at microwave frequencies, so the dissipation at 800 watts of incident power would be tiny so long as there was some air movement to prevent any hot-spots developing.
Not even 5 minutes in the video I'm dreaming of a spectrum analyzer using an "RF prism" in a similar way als old style spectrophotometers use 😁
I have been looking at Luneburg lenses at my previous job for a 5 GHz system but never got to make one or see one working. Following your progress now ia a fun thing to do
It's certainly possible to create a prism for RF, but the dimensions are a bit excessive unless you go to very high frequencies. The problem is that most low loss dielectric materials that are transparent to RF aren't dispersive. The refractive index is pretty much constant with frequency, so your prism can't split out the different wavelengths.
You can do diffraction gratings and grating mirrors as well of course, and they will definitely work at RF.
Dispersing a few sacks of TiO2 or powered alumina in a bathtub or two of molten paraffin wax and setting it in a mould to make a large enough prism would be quite an exercise as well!
Perhaps in a cold climate, you could form a giant ice prism and use that, the RF losses in water ice are fairly reasonable, but even water doesn't have a frequency-dependent refractive index and mmWave/microwave frequencies. You can certainly bend a collimated microwave beam with a really large prism, but it needs to be a few hundred wavelengths across to be useful, so at least 3 x 3 metre faces at 10 GHz. Perhaps as 122 GHz it would be feasible. I'll have to see what is possible. Grating mirrors or transmission gratings, perhaps with a gradient density, should be able to produce a spectral spread sufficient to make a giant mechanical frequency meter using only linear measurements. Hmmmmm.
@@MachiningandMicrowaves Wow, that's a detailed answer I had not expected 👍
You indeed explained the size of lenses at some point in the video and guessed this would be the case. But grating mirrors or transmission gratings might be fun to look at. Ah well, I'm like Archimedes sitting under the tree waiting for some idea to hit my head😂
Maybe that's what the warp coils are... a bunch of focusing lenses, focus em radition into a tight beam and ride it past the speed of light by bump charging it like getting pushed on a swing!
I've always suspected that c is merely an Advisory Speed Limit
@@MachiningandMicrowaves I can't help it, it's what your thumbnail made me thing of on looking at it!
bloody 'ell, mate... rather interesting 20+ minutes... as other have mentioned, one of the first fairly useful explanations of 'apparent velocity' of waves/particles/energy moving through a dielectric.... I'm a bit of a non-engineering engineer, having bootstrapped my way to PhD-level positions, without a single bit of 'paper' (not so easy in DE) and had the good fortune to act in worldwide R&D (almost 3 billion folks have been 'illuminated' by my work...).
Have been following you for a bit, just hit subscribe today, with the 'all' bell..
Other channel??
Okay you can't drop "3 billion people have been illuminated by my work" and not say what it is. If I was to guess, I'd probably go with OLED's though I'm sure I'm interpreting illuminated wrong.
@@ZeLoShady LEDs and illumination, generally, for 80+% of Nokia phones from 2005 to 2012+... 2 billion LEDs per year, 500 million driver ICs, directed development of LED and LED driver technology for 6+ years
Pardon my curiosity. Can the thermal energy emitted by open flames be interpreted as RF noise if the set-up is sensitive enough?
My hand is a good source of microwave energy, I can "hear" the RF noise from it clearly. Flames certainly make enough noise to be detectable. After all, I can hear my trees and the bricks of my neighbour's house perfectly well. With my largest dish antenna, I can hear the thermal noise from the Moon, even though it's pretty damn cold, it's hotter than the deep space background
@@MachiningandMicrowaves Thank you for your reply ,you have been a tremendous help.
As someone who programs both for my job and as a hobby and who is currently in the processing of making my own programming language, I can tell you that the reason why people like programming is because they're different in the head. I hope it's okay, since you didn't link to it in the description, but I copied your matlab program from the video and I'm going to translate it into C since I don't have matlab. I'll have to figure out how that STL function works on my own, but that'll be fun.
That code is a mishmash of other snippets I found and I think there are much better ones out there. It isn't debugged so it probably does weird things at edge cases and probably has holes that need healing. I think I saw an stlwrite C library somewhere, might have been in Qt or something
It was a good video. Pretty cool.
14:51 so far i don't see anything that requires a special printer, the printer described appears to just be a bigger resin printer. Yes the printer has some pumps and a heater to agitate the resin but there are ways to do that with a normal resin printer. And it is a gyroid infill so technically a normal filament printer could do it if the base material won't mess with the functioning of the particles.
Phil explains the reasoning for the use of the Fortify printers at th-cam.com/video/3YMRfw0uWlw/w-d-xo.html
It would be possible to do much of the tech using a hobby printer if the resin was heated and stirred and the film window wasn't damaged by the inclusions, but for production quality results for aerospace applications, I don't think an Elegoo Saturn would be an acceptable choice. How home experiments though, and especially for home-formulated UV resins, perhaps it's feasible. Loads of fun trying though, whatever.
@@MachiningandMicrowaves to be fair aerospace applications for military are far beyond anything anyone else needs just because of the nature of the objects they are trying to detect or avoid being detected by. My guess for the hobby grade stuff is that quality/effectiveness is going to have a large range from useless to above commercial grade in some instances where the individual is really into the stuff and said individual will have modified hobby grade devices to their needs. To be clear those people will likely never reach mill spec aerospace grade stuff but nobody but the military needs to.
@@deltacx1059 Absolutely, and in much of my work I get the same results (or better) than an aerospace manufacturer could, because I have unlimited time and obsession, so I can make stuff that nobody would consider selling or relying on for space comms, warfighting or safety of life applications. Worst that can happen is a shrug of resignation and a small sigh
The speed of light is constant in a vacuum. In a medium, it is affected by the electric fields of the protons and electrons.
"affected" is doing a LOT of work in that sentence. Although solids have a of vacuum within the atoms, the electric fields of the electron clouds are pervasive, and in PTFE or other good dielectrics, the field of the photon induces forces in the molecular bonds which take a little time to respond, but they then generate their own electric field in opposition. A photon isn't affected by an electric field in a vacuum, so why would it be affected by balanced static fields within a solid? When the charged atoms move, that generates an electromagnetic field that superposes with the field of the photon, but the induced EM fields lag slightly as a result of the mass of the atoms and the "springiness" of the atomic bond mediated by the electrons. By the time you've travelled a micrometre or two into the dielectric, the original photon has had so many interactions that most of the energy is now in the little generated fields form all of those interactions. Now of course, all of those interactions create their own fields and those affect all the others. The effect is that by the time the wave emerges from the other side of the dielectric, the superposition now consists of what looks like a delayed photon, despite all of the interactions being at the speed of light.